velocity to prevent water freezing in pipe - subarctic 1

[hrdrok]

Got pointed here from civil guys

Simple (or not),

I need to move ~10,000 usgpm of water over 3000 feet, in subartic conditions, 24/7 run conditions.

Pipe is fully exposed, free draining both ends.

Negligible head (10m +/-) plus line loss.

Heat trace and insulated pipe not an option due to cost.

Mining application

Anybody have a minimum flow velocity to prevent line freezing,Probably sclair DR17, water at low single digits (2-3 degC).

Ambient air temp -20C and warmer.

ANy ideas?

h_r

 

[moltenmetal]

I would imagine your problem would be during start-up or shut-down conditions, which will happen one day. Could you realistically drain this long a line before it froze? On a 3000 ft line, can you guarantee not a single undrained low spot? The sclair pipe might survive a freeze-up, but thawing it would be a b*tch.

Burial below frost line isn't an option? Mining could be anything- as could "sub-arctic"- -20 happens in Southern Ontario routinely in winter, and we're hardly "sub arctic". Is this on rock or soil?

At 10 ft/s, the water (which is near freezing already) is going to be in the pipe for five minutes. Doesn't sound hopeful. Somebody who does snow-making for ski hills might be a useful source of advice.

 

[MrBTU]

Heat trace and insulate the line.

 

[zekeman]

What pipe size?
I assume you mean the minimum flow when you are not in production,flowing the 10,000 gpm or do you mean always flowing the 10,000?
Fairly simple problem
where you have to get the two film coefficients to add up to a thermal resistance over the pipe length.

 

[Compositepro]

I don't see much problem with freezing. 10,000 gpm takes a large pipe. It won't freeze with any significant velocity and will be slow to freeze with zero flow, giving plenty of time to drain the line. That isn't to say there will be no ice formation in the pipe.

 

[zekeman]

Need more info.
pipe size ?
material of pipe
nominal water temp in pipe.

And I think you have no way of fully draining a 3000' run since you are bound to have loops that will be water plugged and will totally freeze there with consequences

 

[desertfox]

hi hrdrok

This file might help.

desertfox

 

[hrdrok]

thanks all for the info

more details

What i am trying to do is size the pipe.

Heat trace and insulation are really not an option due to cost.

The 10,000 usgpm would be continual.

The line can be installed without low points so it will free drain. If it were to ever freeze, we would wait until the thaw and try again the next year.

Most likely a 12/14/16 inch pipe, basically a trade off between line loss due to high velocity and miniizing pumping horsepower, which is 3x90hp for non freezing conditions.

This is just shy of the arctic circle, so of 10 months pumping /year, say march - november, 2-3 of theose months have the potential for -20/-30c conditons.

We curretnly have heat traced low flow lines that freeze even with heat trace.

Our pumps guys have syuggested that 10 m/s is lots of velocity, i need to know how much lower i can go.

Its all about minimzing cap costs without compromising operation due to missed sizing.

tx for the help

 

[TBP]

The chart I have for preventing bare steel water lines from freezing shows (water at 40*F and minus 20*F air temp) shows the following minimum flow rates:

16" line requires 9 GPM/100 feet, 9 X 30 = 270 GPM

14" line requires 8 GPM/100 feet

12" line requires 7 GPM/100 feet

If wind speed is greater than 15 MPH, it recommends doubling the flow rate.

 

[ione]

You can download the freeware program at the link below

http://www.kaimann.de/en/service/kaicalc.html

Amongst the calculation options you can calculate the temperature drop in your line. Input: ambient temperature, line temperature, outer pipe dia, inner pipe dia, insulation thickness (put this value equal to zero), pipe length, mass flow rate.

Change the pipe diameter and consequently the velocity (being given the flow rate) until the temperature is greater then 0 °C.

This is a very very rough approach, anyway…..

 

[ione]

Equation (38) in the attached pdf could be of some help

 

[zekeman]

The standard solution to this problem is
T-T0=(Ti-T0)*e^-(hl/G*cpc
l=3.14*D=3.14*16/12=4
G=Mass flow rate=10000*7.%=75000
hl/G/cp=4/75000=1/19000
starting temp of water, Ti= 3 degC=37 degF
Outside temp, T0 = -20 C= -4 deg f
cp water =1
h = for wind at 30 mph about 20-30/60= .5 BTU/min ft^2 deg b
this is the dominant resistance for metal pipe, plastic will do better
T0=-20C =-4 F
Ti= water temperature at pump
T=-4+(37+4)*e^-1/19000*L
L=3000 ft
T=-4+41*e^-3000/19000=-4+35= 31
I think you have a problem if the pipe is steel and the water is 3 deg C
However if you use plastic pipe with a wall thickness of 1"
the conductance is about
,2/60/1/12=2.4/60=.04BTU/min-ft^2-deg which is about factor of 10 more resistant to heat transfer and is now the dominant resistance term, so the solution with this pipe is,
T=-4+41*e^-300/19000=36 degrees
which looks a lot better
For 2 deg C
T=-4+39*e^-300/19000=34.4 F
DISCLAIMER
There are a number of assumptionsand probably numerical errors here and the results are too close to freezing to give a definitive answer, although we probably are conservative in one respect-- that the heat of fusion requires quite a bit of cooling to start ice formation and also if ice forms it is an additional thermal barrier.

 

[ione]

hrdrok,

Good stuff in the attachment : I think you can find valid guidelines and inputs for your problem.

Take a look at

4-3 Utility distribution systems in frozen ground (ufc_3_130_6.pdf)

12-6 and on (ufc_3_130.pdf).

I hope it helps

 

[ione]

Here the other pdf file