Propane transfer re-visited 1

[rsr32]

Hi,

I need to conduct a product transfer operation but am not sure how to figure out if it will be possible or not. I'm in a bit rusty on my fluid dynamics and looking for some tips on how to approach this.

I have 2 large capacity liquid propane tanks (24,000 gallons) both about 80% full and need to transfer product from one tank to the other. I'll have 100 PSI differential driving the fluid transfer, my question is whether the limit of excess flow valves will be exceeded at that pressure differential.

I've attached a sketch depicting the layout.

what would be the methodology to solve this- I assume solving for velocity at the intake of the discharge tube of tank 1 (given the 100 psid) is the way to go since I can then figure out the flow rate and see if it exceeds the excess flow valve capacity?

Is the D'arcy equation sufficient to solve for the velocity given all the known variables I currently have?

Any help on this would be greatly appreciated,

Thanks in advance,

Dan

 

[GHartmann]

Generally the transfer will occur up to the limit of your excess flow valve = 225 GPM. The pressue drop across the valve overcomes the spring holding the valve open when you exceed 225 GPM.

I would suggest using the Crane 410 book and using the K values etc to calculate your dP. The only caution is to check for flashing of the propane from the higher pressure to the lower pressure resulting in two phase flow (which envolves more complicated formulas)

Just from an experience perspective, the 100 psi will force more flow than 225 GPM (not confirmed by detailed calculations)

How will you generate the 100 psi differential? If both tanks are at ambient temperature of say 80 deg F, they will both have a pressure of approximately 128 psig.

I also assume the level in the "receiving" tank will be nearly empty to receive the contents of the full tank.

Propane tanks are typically not allowed to be filled greater than 85% liquid volume of the tank volume. The 15% "free" volume allows for the pressure variations due to temperature changes of the tank contents.

Sorry if I have pointed out the obvious.

Good luck.

 

[BigInch]

You will need a pump after a little while. GH said, it will be impossible to move anything once the tanks equalize.

"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek

 

[MiketheEngineer]

Don't smoke

 

[rsr32]

Thanks GH, very helpful info.

I do indeed realize the added complexity due to the 2 phase condition present. Basically, from my understanding, this is the process they will be using , I was just asked to determine as a rough estimate and provide a calc, as to whether a 100 psid will create a flow that will exceed or fall below the rated capacity of the excess flow valve. This may sound over simplified when considering all the complexities that exist in the system, unfortunately this was handed down from management personnel who don't have engineering background.

With that, I'm looking to keep it as simple as possible to show that it will work or not


I can get a total head loss value, and after that solve for a velocity but I'm not sure how to correlate that to the pressure differential.

BTW- I believe a compressor will generate the dp

 

[GHartmann]

Lets try it this way:

Start with the 225 GPM as your first flow case

Utilize a simplified Darcy equation to estimate pressure drop per 100 ft of pipe

Utilize your fittings and valves to determine an equivalent length of pipe

Thus for any flow you can calculate your pressure drop

I am pretty sure a 100 psi differential will pass more than 225 GPM for your system.

I do not think you want to put 100 psi additional pressure on your storage tanks (128 + 100 = 228), as it will begin to approach your MAWP (probably 250 - 275 psig). Most transfer systems assume a pressure drop of approximately 30 psi to pass the quantity of liquid you have indicated.

See the attached brochure from Corken showing their propane transfer methods.

Good luck!

 

[BigInch]

There's some good tech publications here.

http://www.corken.com/

"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek

 

[rsr32]

great help, thank you again. The documentation attached is a great resource as well. I will follow your advice and provide them with the recommendations regarding the pressure. It sounds like the flow created with 100 psid will exceed the EFV capacity so I'll have to work my way down to an acceptable pressure level both interms of the EFV and tank MAWP (i believe the relief valves are set in stages 265/275- I'd have to verify though because I'm not familiar with those tanks)

 

[GHartmann]

If you use your flow of 225 GPM and calculate the pressure drop through your system following the formulas in Crane 410, then you can determine the maximum differential allowed. Just informally using a corporate spreadsheet we have, I am getting 35 - 45 psi or so.

Good luck.

 

[BigInch]

Oh. Same website! They do have some good tech info there.

"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek

 

[rsr32]

I will definitely try to get my hands on the Crane 410 book, sounds like they'll have to come way down on their original pressure for this to work..

 

[BigInch]

The Corken manuals have all you need.

"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek

 

[GHartmann]

I didn't think the Corken manual had enough information to determine the "real" piping pressure drop. All of their calculations assume dP of 30 psi for liquid transfer.

Another thing to look at is what type of excess flow valves are installed. Many of them also act as check valves and you won't be able to use the "suction" pipe as an alternate "fill" line.

If you can't find a Crane 410, there are plenty of references on the internet to calculate the dP in your piping.

Again, good luck!

 

[rsr32]

thanks, I've been at it off and on over the course of the afternoon, juggling this and 30 other tasks and by now my brain is fried.. I seem to get stuck once I determined my equivalent pipe lengths on my minor losses and calculating the equivalent pressure drop to determine what my maximum pressure differential can be in order not to exceed the EFV capacity.

As far as the EFV, I was told it would not act as a check valve as they've used the draw tube for filling purposes in the past. Unfortunately, there is no data/specs on the valve itself other than the rated capacity..

 

[rsr32]

here is where I stand at this point.


I used 225gpm and calculated losses in the system based on the velocity at this Q value. It looks to be roughly about 18psi loss across the system. sounds a bit low to me?



Liquid Friction Pressure Loss – 3” pipe sections w/ (2) EFV’s, (2) globe angle valves, (2) 2” reducers

Pressure Loss (psi): 10.14 Head Loss (ft): 47.3

Nominal Pipe Size: 3
Pipe Schedule: SCH 80
Flow Rate (gpm): 225
Viscosity (cP): 0.11
Specific Gravity (water=1): 0.495
Temperature (F): -44
Pipe Roughness (ft): 0.00015
Actual Pipe ID (in.): 2.9
Fluid Velocity (ft/sec): 10.93
Reynolds Number: 1104178
Friction Factor: 0.018
Overall K: 25.49
Piping Length (ft): 22
Lift Check : 2
Angle Valve Flow Down : 2
No. of Reducers: 2
Reducer Outlet Size (in) : 2



Liquid Friction Pressure Loss 2” Pipe section w/ (4) 90deg elbows, (2) 2” globe valves

Pressure Loss (psi): 8.05 Head Loss (ft): 37.6
Nominal Pipe Size: 2
Pipe Schedule: SCH 80
Flow Rate (gpm): 225
Actual Pipe ID (in.): 1.939
Fluid Velocity (ft/sec): 24.46
Reynolds Number: 1651427
Friction Factor: 0.019
Overall K: 4.05
Piping Length (ft): 19.3
Short Radius Elbows: 4
Ball : 2

 

[GHartmann]

You have certainly done the work. Be careful mixing K factors for fittings with equivilent lengths of pipe. Check your literature sources to make sure how to use these properly (sorry its too hard to explain).

You can treat that separately if need be:

dP for straight pipe: Use the dP = f L (RHO) V sq / 2 gc

Then use your K factors for the fittings and apply that formula.

Your K factors look low for the fittings you have.

I agree, with that type of velocity I would think your losses would be higher.

But I did check with a spreadsheet I have and get similar results.


An aside question: How are you keeping the propane in the tanks at - 44 deg F ?

 

[rsr32]

I did look again at the K factors (K= (340)(f_t) as in the case of the angle globe valve) and made sure to plug the value back in to the equivalent lengths formula: Le = (K)(D)/f_t. I 'm double checking my work now but it's been a 13 hour shift so far and the eyes and mind are getting tired. I'll be submitting this tomorrow so I sort of have to finish it now in a way..

as far as dp for the straight pipe, dP = f L (RHO) V sq / 2 gc -- possibly a silly question but what is the 'c' variable stand for?


and with regards to the LPG at -44, this is the value I was given, to be honest I haven't seen them personally as they are out of state and I work in a different department so all I have to go by are figures I was provided. I wish I had seen them for myself because the sketch I got was vague at best. I assume the tanks are possibly vacuum insulated or super-wrap jacketed or perlite blanketed?

 

[GHartmann]

Its g "sub c" or gc the gravitational constant (for those of us that work in English units)

Take a look at the Corken brochure. The propane would have to be at atmospheric pressure to be -44 deg F.

Most propane tanks are unisulated pressurized tanks (128 psig at 80 deg F ambient).

Think of your BBQ grill propane tank, just on a bigger scale.

Anyway you have your conclusion:

A differential pressure of 18 - 20 psi will result in a flow that exceeds the capacity of the excess flow valve.

And you have the calculations to prove that.

I commend you for your efforts and working through the calculations; well done!

 

[rsr32]

I read up on the Corken document, very informative stuff. I did see the pressure/temperature correlation.

I'm personally not very familiar with propane, the only real experience with LPG was as a refrigerant in a LNG liquefaction triple cascade system, and it is sub-cooled in that type of storage tank (lower operating pressure as well)

I've looked at the values I calculated hoping to catch a mistake but no luck, I guess I'll have to accept the result even though my gut feeling would tell me it should be higher than a 18-20 psi drop with a 225 gpm flow through that size piping..

well I do appreciate all your help, it's been a real long day for me and I think that's where I have to call it a night!

 

[BigInch]

If you start with liquid in the line, and a low pressure at the pipe outlet, it will probably flash to vapor at some point along the pipe. After that becomes a two phase flow problem. I don't think that Crane covers those. In fact it is often difficult enough to get LPGs out of a tank and into a pump before it flashes, which is why the Corken vane pumps are good, as they will accept a relatively high % vapor in the inlet stream.

"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek