Diesel fuel flow through a vertical pipe 12

[jreng1]

I am working on a fuel delivery pipe line to deliver diesel fuel to an underground mine. The system is comprised of an open tank on surface with a valve, a 1 ½” vertical pipe line, and an open tank underground.

My question is; Will the flow be uniform and fill the entire pipe to push the air out to the bottom, or will the air tend to make its way up the pipe causing air bubbles.

 

[katmar]

Hi CanuckMiner,

Thanks for the detailed description. I believe the important bit of piping that will determine whether you get gurgling or not, is not the vertical pipe but any horizontal piping between your top tank and where the 1.5" pipe starts its descent. Sounds wrong, I know, but let me explain my thinking.

In a vertical pipe filled with diesel the static head would be 84 kPa per 10 m of height. At equilibrium the diesel fuel would be flowing fast enough down the pipe so that the friction would exactly match the static head providing the driving force. By my calculations this would be roughly 6 kg/s for a 1.5" pipe. This gives a velocity of about 5.5 m/s, which is higher than you would probably normally use in a 1.5" pipe.

Now the head in the top tank has to provide this amount of flow through the tank outlet, plus any valves and the horizontal piping to get to the top of the vertical section. I don't know your configuration, but unless the vertical pipe is virtually directly below the top tank the available head will NOT be enough to drive 6 kg/s through this section of piping.

This will result in the vertical section being fed less than the equilibrium amount of 6 kg/s, and the vertical section will therefore not run full and you could have all sorts of gurgling, and this could potentially lead to dangerous vibrations and hammer. This is the sort of analysis that the paper in Chemical Engineering that I referred to previously deals with.

I think the solution would probably be to use one or two pipe sizes larger for the piping from the top tank to the top of the vertical section. Or make the vertical section one size smaller. Using a 1" pipe for the vertical section would give an equilibrium flow of about 2 kg/s and would drain your 2000 liter tank in about 15 minutes.

Please note that although I have designed and installed a lot of piping over the years this is a new application on me and you should check all these calcs yourself.

regards
Harvey (Katmar)

 

[Montemayor]

jreng1:

With so many good-intentioned replies,it's unfortunate that you don't even mention that your flow is due to gravity; it's also unfortunate that so many responses have been offered but not one has addressed what you describe: a vertical drain line that flows due to gravity. I believe the answer to your basic question (as I understand it) is found at the following website:

http:/

http://www.freecalc.com/selffram.htm

This site calculates a self-venting pipe size - which is the problem you describe. The key word (that hasn't been employed) is: "self-venting". The pipe size must have that characteristic or you will not have steady flow. I seem to remember the need to apply the Froude Number in this type of calculation... nevertheless, the answer is to be found in the above website.

I do not believe you want (or should get) "Plug flow" - or flow that fills the total pipe cross-section. This is the same technique that is used for draining condensers and other process equipment. For a discussion on the subject, read: "Designing Piping for Gravity Flow"; Hill; Chemical Engineering; Sept 05, 1983.

I hope this resolves your question.

Art Montemayor
Spring, TX

 

[katmar]

Hi Art,

I would not design for self-venting in this application for two reasons.

Firstly it would be extremely expensive. The flooded flow through a 1.5" pipe is about 6 kg/s. In order to carry this amount in self-venting mode the pipe would have to be 6". As the underground tank is several thousand feet below the surface this would mean a large cost increase.

Secondly, to be self-venting the pipe must be vertical for its full length. This may not be practical for this length of pipe. I have often installed self-venting piping below surface or direct contact condensers (the application you describe), but it is a very different situation when you have only 8 to 10 meters of vertical pipe to deal with. I would imagine that in a mine it would be very difficult to have several thousand feet of straight pipe with no bends, horizontal or diagonal sections. Jreng1 or CanuckMiner - what would this pipe look like?

The article that you referred to by Hills is probably a lot more accessible to most than the ancient one I mentioned! He also cautions against using bends, although I suspect a typo where he states "gently sloping piping is preferable to vertical runs". I'm sure he meant "gently sloping piping is preferable to HORIZONTAL runs".

I once "inherited" a plant where the drain from a condenser under vacuum had a horizontal section in it, and this plant was very difficult to operate. We changed the piping and it made a huge improvement.

Vertical downflow leads to problems in the intermediate flow regime between self-venting and syphon flow. In my earlier post I worked on the basis of the friction losses exactly balancing the static head, and this would ensure a flowrate higher than the syphon point - thus ensuring that all air bubbles would be swept down with the diesel. Self-venting flow is the other extreme, and it does of course lead to good, safe and operable designs in the correct application. It is only when you try to operate between the syphon and self-venting points that you get gurgling, bubbling, vibration, hammer and air locks.

As I said before, I believe the secret in this application is to ensure that sufficient diesel is fed to the vertical piping to guarantee that it can operate at zero pressure gradient i.e. available head = friction losses.

regards
Harvey (Katmar)

 

[lilliput1]

How about solving for the vertical pipe length that would be self venting at 1 1/2" pipe size. Put valve at each multiple of this length & sequentially open them starting from the top to bottom to deliver the fuel.

 

[25362]

To lilliput1, your advice of ending the filling piping near the bottom of the underground vessel, has it anything to do with static electricity buildup ? Thanks.

 

[Montemayor]

Harvey:

This is a very interesting thread and one that I feel is not only important in its understanding of fluid flow, but also one that deserves more detail and attention as you’ve ably addressed. As I inferred, this query is full of holes that need plugging: no flow rate is stated, no total vertical height is given, and neither is mention made of gravity flow. I think we now all concede that all we have as a driving force is gravity flow; i.e., no pump is being used. Somehow I suspect this is an academic problem, but I totally agree with your description of the basic flow characteristics – except that I don’t quite understand whether you are describing a definitive, derived height of a slug of diesel or if you mean that the entire vertical pipe would be diesel-full

You state that “the diesel fuel would be flowing fast enough down the pipe so that the friction would exactly match the static head providing the driving force” but doesn’t this infer that the entire height is diesel-full? If, as you state, the total height could be “several thousand feet” then we’ve got a hefty design scope that must be designed and made operational:

1) 2,000 vertical feet of diesel fluid yields a hydrostatic head of approximately 700 psig, I believe;
2) If the hydrostatic head pressure at the bottom of the pipe is 700 psig, then an expansion has to be controlled into the target tank using a control valve; this means that flash vapors could be produced and this would have to be taken care of (another subject for a later time).
3) Needless to say, if the pipe is kept diesel full, it must be mechanically designed to withstand the pressure and operate as such.
4) If the total pipe is indeed diesel-full, doesn’t this raise the concern of maintaining such a pressured condition in an underground mine? I believe this concern was mentioned previously.

I agree that a self-venting, 1-1/2” vertical pipe is limited to approximately 5 gpm; but that’s the pipe size that is given. With an assumed specific gravity of 0.8 the 6 kg/sec of diesel is equal to approximately 120 gpm. This seems like a very high flow rate for 1-1/2” pipe and more in line with a 2” pipe size. Could you furnish the equation or method of arriving at the flow of 6 kg/sec?

For operational control of keeping the vertical pipe devoid of diesel when the gravity transfer is not being done, it would seem that self-venting flow is the only way to go. A conceptual design would include a flow control on the outlet of the top source tank prior to its introduction into the vertical gravity fill line. This would allow for shut-off at the source outlet, leaving the vertical line dry after the transfer. I know it was stated that the target tank is smaller than the source; but the query once again fails to state if the method of fill is to transfer 100% of the source tank 100% of the time. I’m assuming that would not necessarily be the case. But then again, I could be guessing wrong.

Regards,


Art Montemayor
Spring, TX

 

[25362]

On a pipe with both ends at atmospheric pressure, an estimated friction drop of 1 ft of liquid per ft of pipe, 120 gpm (18.9 fps) would approximate the maximum flow rate attainable by gravity alone in a vertical pipe, discarding any static head in the above-ground storage tank. If the pipe wouldn't be just vertical but slanted at times, this maximum apparently wouldn't be attainable.

For water (1.13 cS), the same friction drop (in a 1.5" sch 40 pipe) would correspond to a flow rate of 130 gpm (20.5 fps) as tabulated in the Hydraulic Institute Pipe Friction Manual.

 

[Montemayor]

25362:

I agree with your data and references; however, nobody that I have known would design for an average flow velocity of 20 fps in non-compressible flow. That is why I have always used 2" pipe for the range of 120 gpm.

Additionally, although it hasn't been stated (and I believe it is important to note), your basis is that the vertical pipe is 100% full of diesel fluid - which I believe is not accepted in the mininig industry according to CanuckMiner - especially in view of the diesel's high hydrostatic head.

I maintain that the answer to the basic question is that if self-venting pipe is used (at the appropriate flow rate) the flow will be uniform and the air will be equalized between both tanks resulting in NO air bubbles. Agreeably, the pipe is bigger than 1.5" if you desire to transfer at a rate greater than 5 gpm - but that's the inherent tradeoff. If we agree that the scope of work must discard a full vertical line with a control expansion valve at the target tank inlet, then I don't see any other alternative but self-venting pipe.

This continues to be a very important discussion because it involves what seems to be simple, non-compressible flow - but with the caveats that it is vertical flow and deals with a potential flammable and explosive fluid in closed quarters. I think I am correct and logical in my basis, but I am open to other opinions and ideas that can defeat it with a better or safer method.

I also believe you are right in assuming that recommending the filling piping near the bottom of the underground vessel has to do with static electricity buildup. This is an important safety concern when transferring a hydrocarbon into a storage tank. However, I would not fill at the bottom because of obvious fluid driving force reasons. Rather, I would simply drop the gravity flow into the target tank via a dip pipe located in the target tank's roof. I also agree that the vent in the target tank should be directed into a nearby "Return Air Raise". CanuckMiner states "that when the valve is opened, the line "bubbles and gurgles" and drains slowly, then after a period of several minutes the flow picks up and it drains much more rapidly. We are trying to establish the science/physics of what is going on." I strongly believe that the science/physics is readily explained and applied via a self-venting, appropriately designed pipe. I believe the equation that applies is:

D = 0.92 (gpm)^0.4

where D = pipe internal diameter, in.

Note that the height of the pipe doesn't enter into the equation - I don't know why at the moment since I don't have the related paper that explains the equation in my hands at the moment. I am very interested in anyone's explanation or comments on this equation and it's applicability in this example. This subject continues to arouse my interest in other engineers' comments and opinions - such as yours.

Regards




Art Montemayor
Spring, TX

 

[katmar]

Art,

This is the sort of problem that could lead to a really fascinating and productive afternoon for a bunch of engineers grouped around a blackboard. You are correct that there is a lot of missing info, and I have made quite a few unstated assumptions - which is always dangerous! I hope Jreng1 or CanuckMiner can answer the questions in your last post for us.

I had indeed assumed that the entire vertical height would be full of diesel. The way I worked out the flowrate was to calculate the static head per 10 meters of height (not knowing the full height) and then I calculated the flowrate that would give a friction pressure drop over a distance of 10 m equal to this static head. I took the SG as 0.84, giving me a static head of 84 kPa. A standard Moody type friction factor calc gave me a flowrate close to 6 kg/s.

Under these circumstances the friction drop would offset the static head and you would not see the 700 PSI at the bottom of the vertical leg PROVIDED THE DIESEL KEPT FLOWING. If there was a blockage at the bottom the pressure would go to 700 PSI, but this would be true of any design. As you implied, the hydraulics is the easy part of the problem - designing mechanically for a safe installation would be rather tricky.

I had assumed that in each transfer the top tank would be drained completely. If this is not so, then designing for a full vertical leg will not be possible. It would contravene the codes, plus when the line is running there would be something like 3/4 ton of diesel hurtling downwards at 5.5 m/s and closing the top valve might just cause problems!

A 1-1/2" pipe running in self-venting mode with a flowrate of 5 gpm would empty the 2000 liter tank in a bit under two hours. This might be quite adequate if the transfer is only done once per day. However, I am still concernad about designing such a long leg for self-venting flow because there are likely to be some non-vertical sections. The horizontal and sloped sections may have to be made of slightly larger pipe - say 3".

Hi 25362,

Your 120 gpm is VERY close to my 6 kg/s. As you stated, I had assumed pure vertical flow. Once we know what the actual geometry and length of the pipe is, we can match the real pressure drop to the static head and calculate the flowrate. It will be less than what you and I calculated, but unless there is a lot of non-vertical piping the flowrate will still be safely above the syphon point and no air bubbles will rise through the diesel.

regards
Harvey (Katmar)

 

[25362]

http://www.freecalc.com/fluid.htm

gives a maximum of 3 gpm flow for a 1.5" sch 40 vertical pipe to get a self-venting condition. A result that confirms the formula shown by Art Montemayor.

Higher flow rates would imply air would have to be pushed down by the running diesel, and probably heated to some extent.

Art Montemayor would be right in his concern for safety.

This is indeed an interesting subject worthy of thoughtful consideration.

 

[Montemayor]

katmar & 25362:

Thanks for a very stimulating, positive, and constructive analysis of this hydraulic problem. Even if it is merely an academic exercise, we all know that it is any every-day application out in industry and in mines - where the safety concern is correctly addressed. You guys have dissected a query devoid of basic data and presented it in a simplified and correctly defined manner, noting the critical and safety concerns that should be taken into consideration when the application is put into practice.

Harvey:
Your last post gave me a complete and logical understanding of what you have been expounding. I agree completely with your caution that a restriction or block valve at the entrance to the target tank should be PROHIBITED in this application. Additionally, as you've noted, any possible horizontal runs or other potential constraints in the actual field installation should be carefully analyzed. This is something I overlooked in pointing out.

Another point I failed to mention in my haste is the fact that the air vent on the target tank now plays a significant and potentially hazardous role if not designed and sized appropriately. If this vent is undersized, restricted, or worse - blocked off, the target tank can be subjected to huge hydraulic forces that I would prefer not to think about when considering an underground mine.

Summarizing: Your contributions clearly point out what most of us have known by experience - nothing is that simple (especially gravity) that we can assume nothing will go wrong or create a hazard. Simplicity is great, but a knowlegable engineer is still required to design the system and ensure safe operation.


Art Montemayor
Spring, TX

 

[aviat]

I have been around fuel systems like this before, for transferring fuel from the top of a hill down to the bottom, although not to the depth of a mine. When you first start up a system the fuel pushes air out of the hose, first as a slug then bubbles then some fluid, then you get another air pocket and it starts over again. This continues until the system is bleed. In your case the air would go into the bottom tank and then out it’s vent. Once the air is out of the system it will continue to flow until you run out of fuel.

As for why the fuel flows slowly at first and then flows faster after. The top tank has atmospheric pressure acting down on the fuel, when you open the valve the weight of the fuel pushes down but both of these are opposed by a higher air pressure in the mine and the resistance of the piping. Once the weight of the fuel pushes the air out of the pipe, it then acts along with the air pressure to push the fuel down the pipe. This is a common problem with gravity systems, even with systems closer to the same level. Sometimes you cannot get system to flow at all, but there is a simple solution. You just add air to the top of the top tank to increase its pressure and it will force fluid down the lines.

In your system you can control flow by closing the valve in the exit from tank. According to ASTM D 4865 05.02 Standard Guide for Generation of Static Electricity in Petroleum Fuel Systems you should limit velocities to below 7 m/s. This standard recommends bonding the system. If you require info on that I can provide that. I would recommend installing a hose larger than the piping, between the pipe and the tank at the bottom, to allow the air to exit. The hose expands and lets air though better than a solid pipe.

Good luck!

 

[bulkhandling]

I'm trying to draw a picture about the "bubbles and gurgles" phenomenon as described by CanuckMiner.

Valve is opened, fluid flows into the verticle pipe. fluid does not fill all the pipe section area but flows along part of the pipe wall. The very front fluid flows slower because it has to overcome surface energy of Solid-Air to form the surfaces of Solid-Liquid, Liquid-Air. The after fluid flows faster and jioned the very front fluid. Eventually it formed a full area fluid column and enclosed air above it that I call it air POCKET. The air pocket blocks the flow just as solids, so the fluid flows slowly. The air pocket is pushed up by the fluid similar to the air bubbles floating up in a tank. When all the air is pushed out thru the valve, the real fluid column formed that follows the formula of Art Montemayor. So you get the regular flowrate now. You hear gurgling because fluid flows in the pipe with air in it.

The higher air pressure in the underground slows the flow but not significent. The pressure is only 1.25psi higher than the above ground atm pressure and it never disappears even after all the air is out of the pipe. It is fairy small comparing to the static head.

 

[aviat]

I found a report describing how they position ice into deep mines in Africa. It describes some things I did not in my post last week. Cohesive and adhesive forces between piping and fluid restrict the fluid in the piping. In the vertical runs the fluid overtakes the air, but the air is first compressed, and exits from the pipe under pressure. When fluid overtakes the air it reaches equilibrium and only goes so fast. Also at all bends air bubbles are created and they are swept down the pipe.

As for the sounds I think it would depend where you are. If you are standing near the vertical when the air is being overtaken you would hear burbling and if you were there when flowing freely it would sound like running water similar to a sewer pipe in a tall building. If you are at the pump outlet you get have a pipe, which is open at one end and closed at the other. This creates standing waves that come down the pipe and exit then return back up the pipe. This is covered in physics textbooks. The bottom line is that after a few times you can tell by the frequency of the sound when the fluid will be coming out.

A lot of this applies when you first start a system with a pump. The fluid will overtake the air and come out in spurts of air and fluid.

As far as a formula to determine flow rate you would have to wait until all the air is out of system and are achieving steady flow. Without knowing the amount of the vertical and horizontal runs and number of bends it would very hard to come up with a number. The easiest way to measure flow with system operating in this case would be to have a known volume and measure the time for it to go. Time should be between 5-10 minutes and then average it for flow per minute.

http://deepmine.csir.co.za/Publications/Chapter 52.pdf

http://www.ualberta.ca/~jplambec/che/p102/p02306.htm

Happy New Year!

 

[lilliput1]

Consider this proposed solution.
Put hand pump at top of top tank with suction line droping down to tank. at top of discharge line above pump put valved funnel for use for either priming pump or venting the supply pipe to empty it. Put valved bypass around the pump suction & discharge, upstream of the valved funnel. Tee off to side then drop the supply piping upstream of the connection to the valved funnel. To supply oil: close bypass valve; open funnel to prime pump then close; operate the hand pump to push oil & air down the supply piping; whn oil is flowing steady to bottom tank, open bypass & let flow to be by syphon; after oil has been transfered, open funnel to break the syphon & vent the vertical supply line.

 

[GenB]

In answering to the first posting:
Air will travel up the filling pipe if you drop the fuel at the top empty part of the underground tk.
if you fill at the bottom, whe the underground still has some fuel in it, there will be no air traveling up the fuel filling pipe.
Provided you have a vent ion the underground, everything will work well.

Now if we talk about the vent: if it is impossible to vent by installing a second pipe, I would insert a smaller (maybe a poly line or a flex matl compatible w/the fuel) inside the filing pipe up and above the above-ground tank.

ER