Forums
Blogs
Articles
File Library
FAQ
Search
FB
Does anyone know of a reference, article, or rule of thumb for sizing a line so that fluid flowing vertically downward inside will keep it full? For instance, I have a pumparound loop, going into the head space of a tank. Given a flowrate and pipe diam, am trying to determine if the pipe is liquid full, or if liquid is running down just the inside wall of the pipe. This will affect discharge head calculations, seal loop design, etc.
I've heard a rule of thumb is 5 linear ft/sec, but am inquiring about the basis and background. (What happens when density or viscosity change greatly, etc.) Nothing apparent in Perry's, nothing in Crane...unless I missed something obvious.
Any help is appreciated.
Thanks!
|
|
Pipe Flow
Started by Guest_Anonymous_*, Jun 19 2003 04:12 PM
6 replies to this topic
Share this topic:
#2
-
- guestGuests
- 0 posts
Posted 24 June 2003 - 07:44 AM
Hi. Ive seen a velocity table in *Rules of Thumb for Chemical Engineers" . However, it only states the recommended velocity for the type of fluid. For water, the recommended velocity is 5-7ft/sec.
I dont think Ive seen table of velocity at choking condition.
I dont think Ive seen table of velocity at choking condition.
#3
Guest_Guest_Art Montemayor_*
Guest_Guest_Art Montemayor_*
-
- guestGuests
Posted 24 June 2003 - 09:30 PM
Anonymous:
What you describe has very little to do with the potential resistance at the end of your "pump-around" loop. If you have a conventional pump-around it is made up of :
1) the horizontal pump suction line;
2) the vertical pump discharge line;
3) the horizontal pump discharge line at the top of the tank; and,
4) the vertical pump discharge line at the top of the tank.
I realize that the above is simplified, but my point is that section (4) is what you show any concern about and this is normally a very short section run. Besides, even if that section runs half or partly full, the resistance to flow will be less than if it were 100% full. So, if you assume that that section is running full of liquid, your resistance for the entire loop will be conservatively high in the calculations. This is the way it normally is calculated - unless you have a significantly long vertical, downflow and you MUST have an accurate answer. This is normally not the case but if it were, then you must apply the Froude number to your flow analysis and calculations.
The Froude number is what is used to determine self-venting in downward flow of liquids. This becomes important, for example, when you need to have an overheads condenser drain its liquid product freely through a vertical down-flowing drain pipe. The pipe must have self-venting capabilities, or it will not drain efficiently.
The liquid velocity of 5-7 fps that is being mentioned is for a 100% liquid-full system. The velocity is not what sets the ability to have successful flow. You can have greater velocities than that -- if you can supply more available pressure drop. There is nothing magical about the fluid velocity that makes a system work or not work. The system resistance to flow is what allows it to work and it depends on velocity and other factors as well.
If this is a normal industrial application, I wouldn't worry about the partially-full portion in the last discharge section into the tank. To calculate the horsepower of your pump, assume a 100% liquid-full system and rely on a discharge throttle valve to control the flow on the centrifugal pump (assuming that is the type you use). If this is an academic problem, then you have to do your homework on the Froude number.
Also, let me make the point that engineers shouldn't toss around jargon without being knowledgeable about its meaning or understanding. This causes confusion and erroneous communications. It impresses no one. The use of the term "choking condition" has no meaning or value in this conversation. This is described as a LIQUID system and choking conditions in Fluid Mechanics applies to GASEOUS systems. Choked flow is something totally different and foreign to the described system. Petroleum engineers sometimes use the term "choked" to describe a throttled liquid (as through a valve). But they can be forgiven since they control the environment and the language out in the Oil Patch, where no other engineer desires to go.
Art Montemayor
What you describe has very little to do with the potential resistance at the end of your "pump-around" loop. If you have a conventional pump-around it is made up of :
1) the horizontal pump suction line;
2) the vertical pump discharge line;
3) the horizontal pump discharge line at the top of the tank; and,
4) the vertical pump discharge line at the top of the tank.
I realize that the above is simplified, but my point is that section (4) is what you show any concern about and this is normally a very short section run. Besides, even if that section runs half or partly full, the resistance to flow will be less than if it were 100% full. So, if you assume that that section is running full of liquid, your resistance for the entire loop will be conservatively high in the calculations. This is the way it normally is calculated - unless you have a significantly long vertical, downflow and you MUST have an accurate answer. This is normally not the case but if it were, then you must apply the Froude number to your flow analysis and calculations.
The Froude number is what is used to determine self-venting in downward flow of liquids. This becomes important, for example, when you need to have an overheads condenser drain its liquid product freely through a vertical down-flowing drain pipe. The pipe must have self-venting capabilities, or it will not drain efficiently.
The liquid velocity of 5-7 fps that is being mentioned is for a 100% liquid-full system. The velocity is not what sets the ability to have successful flow. You can have greater velocities than that -- if you can supply more available pressure drop. There is nothing magical about the fluid velocity that makes a system work or not work. The system resistance to flow is what allows it to work and it depends on velocity and other factors as well.
If this is a normal industrial application, I wouldn't worry about the partially-full portion in the last discharge section into the tank. To calculate the horsepower of your pump, assume a 100% liquid-full system and rely on a discharge throttle valve to control the flow on the centrifugal pump (assuming that is the type you use). If this is an academic problem, then you have to do your homework on the Froude number.
Also, let me make the point that engineers shouldn't toss around jargon without being knowledgeable about its meaning or understanding. This causes confusion and erroneous communications. It impresses no one. The use of the term "choking condition" has no meaning or value in this conversation. This is described as a LIQUID system and choking conditions in Fluid Mechanics applies to GASEOUS systems. Choked flow is something totally different and foreign to the described system. Petroleum engineers sometimes use the term "choked" to describe a throttled liquid (as through a valve). But they can be forgiven since they control the environment and the language out in the Oil Patch, where no other engineer desires to go.
Art Montemayor
#4
Guest_Anonymous_*
Guest_Anonymous_*
-
- guestGuests
Posted 25 June 2003 - 12:47 PM
Art:
Thanks for your reply, it is very helpful.
Our return line to the top of the vessel is long.
My concern is with the total discharge head (I'm using Durco's pump engineering manual, Edn 5) which is comprised of friction head (minimal - even with longer pipe run, as you mentioned), discharge surface pressure, and static head. In example 3 in this manual, the static head is calculated by subtracting the height of the vertically down run from the height of the vertically up run. (In parentheses, the manual also states that this value is "based on the assumption that the vertical leg in the discharge tank is full of liquid and that as this liquid falls, it pulls the liquid up and over the loop in the pipe...normally called a siphon leg"). This goes back to my original question, which is addressed with the Froude Number.
I will assume both cases (liquid line full versus liquid line self-drained) for the discharge head and see what I get. The difference in height is not insignificant when looking at the pump's TDH.
Any more advice on the Froude number?
Thanks for your reply, it is very helpful.
Our return line to the top of the vessel is long.
My concern is with the total discharge head (I'm using Durco's pump engineering manual, Edn 5) which is comprised of friction head (minimal - even with longer pipe run, as you mentioned), discharge surface pressure, and static head. In example 3 in this manual, the static head is calculated by subtracting the height of the vertically down run from the height of the vertically up run. (In parentheses, the manual also states that this value is "based on the assumption that the vertical leg in the discharge tank is full of liquid and that as this liquid falls, it pulls the liquid up and over the loop in the pipe...normally called a siphon leg"). This goes back to my original question, which is addressed with the Froude Number.
I will assume both cases (liquid line full versus liquid line self-drained) for the discharge head and see what I get. The difference in height is not insignificant when looking at the pump's TDH.
Any more advice on the Froude number?
#5
Guest_Guest_*
Guest_Guest_*
-
- guestGuests
Posted 27 June 2003 - 11:07 PM
Anonymous:
Now that you've given more detail and cited the Durco Engineering Manual (5th Edition), I believe I see what is hindering your understanding of the Total System Head.
I also have a well-worn 5th Ed. of the old Durco Engineering Manual and my copy shows my liberal red marks from years past. On page 47, I put a large red comment on the first paragraph: "This is an Optimistic assumption; the best and safest practical engineering design criteria is to assume zero recovery." I remember now these comments and recall that what was discussed in Example 3 is the assumption that if a vertically, downflowing pipe is 100% full you will "recover" the energy that it took to elevate it. In this example the drop of the vertical leg is 10 feet. The term I have always heard and used is "total (100%) recovery", and I was warned by my mentors many years ago that it was foolhardy to believe that you could depend on recovering the hydrostatic head represented by that discharge vertical drop. I have never assumed this in 43 years of engineering and subsequently never had any problems with my applications. I still stand by what I originally recommended: don't depend on recovering this energy. Be conservative in your design on this issue when designing pumping loops. I had the opportunity to discuss this issue with Durco engineers in 1990 and they agreed that my philosophy was sound and this portion of their manual was being academic and theoretical. Note that I am referring to example 3 and NOT example 4. Example 4 is a totally different hydraulic situation because in this case you don't have an "air break".
Allow me to make a comment on the old Durco Engineering Manual. This excellent book is now the FlowServe Pump Engineering Manual and is still available free of charge from FlowServe Representatives. I strongly recommend any young engineer to solicit and obtain a copy as soon as he/she are able to and study it religiously. You will never regret spending valuable time on it. This is one of the best pump primer books ever published, but it has its flaws and errata. For example, note:
1) Page 35 erroneously states that the Hydraulic Institute's Pipe Friction Manual has charts that allow you to read the friction factor directly. This is wrong. The friction LOSS (hf) is read directly in given TABLES;
2) Page 38 states that conversions are provided in Table 3 in the Appendix. This no longer applies; the table was changed from what it was in the 5th Edition;
3) Page 57 states that Table 4 in the Appendix contains equivalent lengths. This Table 4 was changed in the 6th Ed. and equivalent length values are not to be found in the 6th Ed.
4) Throughout the manual, the Hydraulic Institute's Figures are erroneously identified with consistency.
There are a few more errata, but they don't deter from the fact that this manual is of great value to engineers. In fact, I learned a lot about hydraulics by finding the errata in this manual.
From what you've described as the system, I would not apply the Froude Number to prove that you have/don't have a 100% liquid drop leg. Just go with what I'm suggesting and forget about 100% recovery.
Now that you've given more detail and cited the Durco Engineering Manual (5th Edition), I believe I see what is hindering your understanding of the Total System Head.
I also have a well-worn 5th Ed. of the old Durco Engineering Manual and my copy shows my liberal red marks from years past. On page 47, I put a large red comment on the first paragraph: "This is an Optimistic assumption; the best and safest practical engineering design criteria is to assume zero recovery." I remember now these comments and recall that what was discussed in Example 3 is the assumption that if a vertically, downflowing pipe is 100% full you will "recover" the energy that it took to elevate it. In this example the drop of the vertical leg is 10 feet. The term I have always heard and used is "total (100%) recovery", and I was warned by my mentors many years ago that it was foolhardy to believe that you could depend on recovering the hydrostatic head represented by that discharge vertical drop. I have never assumed this in 43 years of engineering and subsequently never had any problems with my applications. I still stand by what I originally recommended: don't depend on recovering this energy. Be conservative in your design on this issue when designing pumping loops. I had the opportunity to discuss this issue with Durco engineers in 1990 and they agreed that my philosophy was sound and this portion of their manual was being academic and theoretical. Note that I am referring to example 3 and NOT example 4. Example 4 is a totally different hydraulic situation because in this case you don't have an "air break".
Allow me to make a comment on the old Durco Engineering Manual. This excellent book is now the FlowServe Pump Engineering Manual and is still available free of charge from FlowServe Representatives. I strongly recommend any young engineer to solicit and obtain a copy as soon as he/she are able to and study it religiously. You will never regret spending valuable time on it. This is one of the best pump primer books ever published, but it has its flaws and errata. For example, note:
1) Page 35 erroneously states that the Hydraulic Institute's Pipe Friction Manual has charts that allow you to read the friction factor directly. This is wrong. The friction LOSS (hf) is read directly in given TABLES;
2) Page 38 states that conversions are provided in Table 3 in the Appendix. This no longer applies; the table was changed from what it was in the 5th Edition;
3) Page 57 states that Table 4 in the Appendix contains equivalent lengths. This Table 4 was changed in the 6th Ed. and equivalent length values are not to be found in the 6th Ed.
4) Throughout the manual, the Hydraulic Institute's Figures are erroneously identified with consistency.
There are a few more errata, but they don't deter from the fact that this manual is of great value to engineers. In fact, I learned a lot about hydraulics by finding the errata in this manual.
From what you've described as the system, I would not apply the Froude Number to prove that you have/don't have a 100% liquid drop leg. Just go with what I'm suggesting and forget about 100% recovery.
#6
Guest_Anonymous_*
Guest_Anonymous_*
-
- guestGuests
Posted 30 June 2003 - 12:22 PM
Dear Guest:
Thanks much for the advice, it looks like it was well worth the question being asked.
I see that there have been quite a few people reading this particular post...I hope the responses have been enlightening to those who have been reading.
Thanks much for the advice, it looks like it was well worth the question being asked.
I see that there have been quite a few people reading this particular post...I hope the responses have been enlightening to those who have been reading.
#7
hasani
-
- Members
-
- 6 posts
Brand New Member
Posted 01 July 2003 - 09:37 AM
Hi everyone..especially Art montemayor...!!!!
Its nice to see u here too...art..Last time was in eng-tips, I must say that your advise on a topic is really helpfull,but I would like to contact u too,can u send me your e-mail address...if u dont mind!.
hasani..u can send me an e-mail via this forum!
looking forward for a reply from your side.
bye
Its nice to see u here too...art..Last time was in eng-tips, I must say that your advise on a topic is really helpfull,but I would like to contact u too,can u send me your e-mail address...if u dont mind!.
hasani..u can send me an e-mail via this forum!
looking forward for a reply from your side.
bye
Similar Topics
Heat Exchanger Steam FlowStarted by Guest_aliebrahem17_* , 25 Nov 2024 |
|
|
||
Flow Through Normally No Flow LineStarted by Guest_iippure_* , 08 Apr 2025 |
|
|
||
Flare Header Reverse FlowStarted by Guest_Ahmadhamzahperta_* , 04 Apr 2025 |
|
|
||
Dynamic Simulation After Feed Flow ReductionStarted by Guest_Kakashi-01_* , 20 Mar 2025 |
|
|
||
Smr Reformer Flue Gas FlowStarted by Guest_kaidlut_* , 14 Feb 2025 |
|
|
