Critique of Mike Soltys Hose Length Calculator

[LayerUp]

Per the request of several active members of the HBT kegging forums, I have upload a derivation of the proper analysis of hose length, as a dissenting opinion to Mike Soltys calculator (for reference, here is a link to Mike's calculator).
http://www.mikesoltys.com/2012/09/17/determining-proper-hose-length-for-your-kegerator/

I attempted to post a reply on Mike's web page around a month ago, and as my piece has not yet been published to the site, either the site is currently abandoned, or the author has no interest in addressing shortcomings in his methods. In the attached .pdf, I have provide some insight, as well as a simple, and complete, derivation for how to calculate the required hose length in a closed channel fluid system.

After reading through Mike's calculation, 3 issues immediately came to my attention. First, Mike fails to consider the energy within the fluid as it leaves the faucet; Mike does not address the velocity term at the faucet. Second, Mike makes a bold assumption that all of the energy loss within the system is due to pipe friction, rather than a combination of pipe fittings and pipe friction, I summed this up in the previous thread as:
"Second, and more damning, is that he assumes that all the energy loss in the system comes from friction within the tubing, rather than the connections within the draft system. This tells me the author has a poor understanding of how closed channel fluid systems operate. He probably read that friction losses in the pipe are called "major losses", and that losses in the fittings are called "minor losses". In many fluids problems, that is correct, but in many fluids problems, we are dealing with 500+ feet of tubing, not 10 feet"
Third, Mike assumes that the user is able to adjust the flow velocity, now that may be true for those who have flow control functionality, many users (myself included), do not have this capability, so the pour speed is based on the physical parameters of the kegging system, rather than a chosen flow rate.

If there is sufficient interest, I may be able to write an executable that I can host on a 3rd part site that is based off of my analysis, and would allow the user to compare to Mike's solution, and the hose length that works for their system. As I do not wish to purchase and maintain a domain for hosting the site, it would likely be a download executable, with the source code, rather than a web based application.

Please excuse the handwriting, I realize it is anything but easy to read, I tried to comment everything out in the .pdf to ease reading....

I am open to feed back, if there are any questions or concerns, or if you believe you have found an error or incorrect consideration, please let me know...

 

[imasickboy]

So, as best as I can tell, Mike's calculator says to use a longer line than is required in your math?

I respect all the effort, but that's not a bad thing. Too short is a bad thing.

And since you don't have a calculator hosted online, Mike's is the best we have available to us. It's up to you to change that. Until then, people will still point others to Mike's tool to answer their line length questions. No one will say "Here's this sloppy pdf, read it and try to do the math yourself."

I give you a solid A for effort, but if you want to be the guy, you've got to finish the job. I'm looking forward to seeing you, (or someone else), put a calculator using your theory out there for real world testing.

 

[ajdelange]

Third, Mike assumes that the user is able to adjust the flow velocity, now that may be true for those who have flow control functionality, many users (myself included), do not have this capability,

I think most of us know how to rig a 'choker'.

 

[mabrungard]

The discharge valve (tap) is a 'minor' loss that can vary from very little to infinity. Its the throttling of flow with the tap that has the potential to create most foaming. Only under full (or mostly full) flow does the 'major' loss of pipe friction aid in creating a low gradient reduction in discharge pressure which creates a nice pour.

 

[ajdelange]

The idea is that with the optimum flow rate (8 sec to fill a 12 Oz glass --> 1.5 Oz/sec IIRK) the friction in the line is such that there is a smooth gradient between keg pressure at the keg end and atmospheric pressure at the dispense end. Thus the pressure across the valve is very low - just high enough to give 1.5 Oz/sec (if that's the right number) at the valve's Cv. I think that may be what you are trying to say.

 

[LayerUp]

So, as best as I can tell, Mike's calculator says to use a longer line than is required in your math?

I respect all the effort, but that's not a bad thing. Too short is a bad thing.

Not sure I follow this argument. My issue is that folk seem to preach Mike's answer as the gospel, that it is THE solution to the serving tubing question. I am simply offering a viewpoint as an addendum to his approach, and showing that his method does not fully consider the interactions within the closed channel flow field.

that's not a bad thing. Too short is a bad thing.

If that is your thought, why use Mike's calculator to begin with? If you don't care about too much line being sized, why not just install 15 feet as the initial sizing on every single job, and adjust the line length as necessary. While that approach is fine in it's own right (hell, it's what I did, until I sat down and solved the problem), but my concern was more for the new home kegerator operator (of which I am), who may wonder why his/her system does not operate as expected, when using this calculator. If this new operator stumbles upon his calculator, the operator may wonder why their system is not reacting appropriately, rather than considering that perhaps the calculation is incorrect.

 

[LayerUp]

I think most of us know how to rig a 'choker'.

Count me as one who doesn't. What is a 'choker', is it some sort of piping reduction? Is it a flow controller? How does that address what you have quoted? My argument was that I expected many to not have a mechanism to fully control the flow rate. Are you telling me that this is an incorrect assumption? If so, that is fine, but I would enjoy seeing some details that indicate this.

 

[passedpawn]

I'm assuming a choker is a reduced ID near the tap, including flow-control taps. I have four of these.

I think the best possible solution is to have a continuous inner diameter that drops the pressure such that you get the perfect flow rate into the glass. Chokers, in my experience, add turbulence that causes foaming.

So, instead of a choker (flow control at the tap), I'd support the math that determines the ideal length without the choker.

 

[LayerUp]

The discharge valve (tap) is a 'minor' loss that can vary from very little to infinity. Its the throttling of flow with the tap that has the potential to create most foaming.

Can you elaborate on the verb 'throttling'?

Only under full (or mostly full) flow does the 'major' loss of pipe friction aid in creating a low gradient reduction in discharge pressure which creates a nice pour.

Fluid flow does not need to fill a pipe to experience friction and turbulent losses in energy. If the pipe is not full, you are simply swapping out the Darcy-Weisbach solution with the Manning-Gaucker formula for energy loss. If the pipe is not full, however, it cannot be pressurized, so the only loss in energy that can occur is that which is due to elevation change, or due to a change in flow velocity, if the hydraulic radius changes.

 

[ajdelange]

Count me as one who doesn't. What is a 'choker', is it some sort of piping reduction? Is it a flow controller?

If one wants to maintain a certain keg pressure (in order to have a desired number of volumes) and experiences an excessive flow rate with the length of hose he has he simply slips in a bit of additional hose to provide the extra friction necessary to slow the flow to the desired value. Many people with draft setups keep a couple of pieces of coiled up hose (in which sense they resemble the 'chokes' of electronics) with tail pieces installed and a supply of 'bullets' (nipples) handy for this purpose. At least I do. That's a choker.

 

[day_trippr]

[...]My issue is that folk seem to preach Mike's answer as the gospel, that it is THE solution to the serving tubing question.[...]

See, it's your perception that's the problem. Given that the other 99.9% of purported line length calculators all start with a totally bogus assumption wrt line resistance, it's pretty darned easy to call Mike's calculator "The Only Line Length Calculator Worth Using".

It's really that simple...

Cheers!

 

[ajdelange]

If the pipe is not full, however, it cannot be pressurized,

One place we frequently see a pressurized pipe that is not full in is the liquid line in a refrigeration system that is not fully charged. In fact that's what the site glass is there for. If you see the line isn't full it's time to look for a leak, fix it and top off the charge.

 

[day_trippr]

Very diplomatic.
Certainly a beer line that has accumulated CO2 bubbles due to out-gassing is still under pressure...

Cheers!

 

[ajdelange]

Very diplomatic.
Certainly a beer line that has accumulated CO2 bubbles due to out-gassing is still under pressure...

Funny I didn't think of that. Probably because the majority of the time I have recently spent in and around my beer cooler was spent on refrigeration lines, not beer lines.

 

[mabrungard]

Can you elaborate on the verb 'throttling'?

Sure, its the control of flow (or an engine...or internet usage) with a throttle. In the case of beer flow, partial opening of the beer tap.

Fluid flow does not need to fill a pipe to experience friction and turbulent losses in energy. If the pipe is not full, you are simply swapping out the Darcy-Weisbach solution with the Manning-Gaucker formula for energy loss. If the pipe is not full, however, it cannot be pressurized, so the only loss in energy that can occur is that which is due to elevation change, or due to a change in flow velocity, if the hydraulic radius changes.

As pointed out above, you forgot the case in which air or gas is entrained in the pipeline. I just finished the design of a 1.5 mile long pipeline that had hills and valleys in the route and I had to confirm if air would be trapped in the line or be flushed along with the flow. Thankfully, the adverse slopes weren't too steep to prevent air from being flushed out with the water flow.

 

[LayerUp]

One place we frequently see a pressurized pipe that is not full in is the liquid line in a refrigeration system that is not fully charged. In fact that's what the site glass is there for. If you see the line isn't full it's time to look for a leak, fix it and top off the charge.

Fell off the face of the earth on this website for while, but sorry boys, I'm back.

To your point here, yes, the liquid line that is not fully charged will be pressurized, that is due to the vapor pressure of the refrigerant, and the closed state of the conduit carrying the refrigerant. Punch a hole in the refrigerant line and expose it to the atmosphere, and the system will not remain pressurized for very long. This what I was trying to describe, albeit, rather unsuccessfully.

 

[LayerUp]

Is there decent file hosting service that is normally used for this website, that I can link to? Per Day_trippr's initial request, I wrote out the analytical solution as a locally installed executable, but I have to admit I'm not sure of the best way to distribute it. Can anyone point me to a recommended file host? Thanks,

 

[LayerUp]

Edit: nevermind, I think I can simply attach it to this thread directly.

 

[Gnomebrewer]

Am I the only one that finds beer line calculators pointless? There are so many little things that can affect your pour that surely it's easier to start with a line much longer than you need, and pull a beer. Too slow? Cut some off. Pull another beer. Too slow? Cut some more off. Etc, until you get the perfect line length for your system. You might waste a dollar or two worth of hose to get it right.

 

[ajdelange]

That's what the aforementioned chokers are for.

 

[shoengine]

I start with a long line and adjust the vertical difference if it is flowing too quickly.

 

[ajdelange]

Curious as to how you do that. Do you keep your kegs on the stairs? On adjustable stillage?

 

[shoengine]

I don't have a kegging system, per se, just a length of hose and a picnic tap. I'm a keg newbie but I read that the velocity of the fluid is due to the headspace pressure but that restrictions such as the elevation that it has to climb and the size, temperature and length of hose all slow it down to reduce foaming. So if my pressure is too high in a particular beer for my 10' line, I raise the picnic tap up higher and it helps reduce foaming.

Note that this method is really stupid, just something I noticed. I use my taps mainly for sampling before bottling from a keg, so I don't really invest much time and expense into setting up a robust system. I just thought it was interesting enough to post. Don't let me derail the thread though, there is some good critique here.

 

[day_trippr]

hahaha! That's a great answer to what otherwise was rather mystifying

Cheers!

 

[day_trippr]

I tried the above linked calculator. Where Mike's calculator recommends roughly .9 feet of 3/16" ID solid pvc tubing per psi, the challenger recommends ~ .6'.
fwiw, my system runs 1 foot per psi and performs very well winter through summer, keg after keg. I wouldn't go shorter...

Cheers!