Internal Natural Convection in a horizontal pipe containing Jet A 1

[MRMaixner]

Looking for a correlation for internal heat transfer for a horizontal circular pipe experiencing natural convection. The pipe contains jet fuel.

Please email at michael.maixner@usafa.edu

 

[fegenbush]

More information will be required to understand fully what you are asking.

Is the pipe the hot or cold fluid?
Diameter and wall thickness of the pipe?
Is the fluid flowing in the pipe?

 

[MRMaixner]

The line is a 12" OD--don't know the wall thickness at this point. Line contains hot JP8, and I'm trying to perform a heat transfer calculation on it as it releases heat to the surrounding soil--the centerline of the pipe is 55" below ground level. The fluid is undergoing a hydrostatic test, and as it cools, the pressure drops, giving a false positive indication of a leak. Am trying to calculate the allowable pressure drop, beyond which the likelihood of a real leak is higher.

Am also interested in calculating the heat transfer while the fluid is flowing.

Many thanks

 

[IRstuff]

The density of the fuel is changing, so pressure should not be used as a measure of quantity:

http://wiki.answers.com/Q/How_does_temperature_affect_the_density_of_gasoline_or_petrol

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[srfish]

Hi,

One reference for calculating the natural convection heat transfer coefficient inside a pipe can be found in good old Kern. it depends upon a variable called the Grashof Number.

 

[ione]

Why not put a temperature probe in the pipe and measure the temperature profile during the test. For each specific temperature you can calculate the pressure variation due to cooling

http://www.eng-tips.com/faqs.cfm?fid=1339

This should help you to establish the threshold you're looking for

 

[racookpe1978]

If you have no flow in the pipe - since you have a pressure test going on! you "should" have no flow - then the temp's will change with ground cover and ground conditions and time of day of the test and depth of burial of the pipe. Go deep enough, and there will be no change worth mentioning.


Example: You go under a runway, the thick hot sun-heated concrete above the pipe will cause that 200 feet under the runway to be changing temperature differently than the 1/4 mile going towards the runway under grass and (thin) asphalt from the tank towards the runway.

Get reasonable temperature readings of the pipe at reasonable intervals down its length.

But I'm making assumptions! I don't know enough about your conditions to really know.

Did you recently make a repair that you are trying to test with a pressure test? Is there massive leakage from unknown places you suspect? Could there be theft and pillage from people draining the tank that they are trying to cover up as a "leak in the pipe"?

 

[MRMaixner]

racookpe1978--thanks. In fact, the temperature at a depth of 20 ft is about 69F year-round--near Las Vegas. The centerline of the pipe is at 55 inches, so that it experiences quite a bit of fluctuation depending on time of year, time of day/night, etc. The pipe is still going to be transferring heat out, since the fuel is (nowadays...) 95F; the thermodynamic properties of the fuel @ 130 psig (compressed liquid region)are such that even a small temperature drop(couple of degrees) can result in a substantial drop in pressure, which would give operators a false positive indication of a (potentially massive) leak--lots of manhours are expended in verifying system integrity, only to find that there was no leak--this is on a USAF base, so no one is pilfering the fuel--it's a 3000 ft run between the tank farm and the airfield. Over the period of the 1-hour hydro, 36 psi drops have been experienced. Current guidelines specify that the max allowable pressure drop is 4 psi--this might be OK when you're performing a hydro with water, but NOT with something that has the characteristics of JP8/JetA--they're using the fuel that's normally pumped as the hydrostatic test fluid, and can't afford the time to drain, fill with water, hydro, flush, refill, etc.

Bottom line: In order to provide operators with a revised pressure drop table based on ambient temperature data (I'm recommending that a thermocouple be placed down at the pipe--they already measure the fuel temperature before isolating the system). I still need a guideline for Rayleigh/Grashof number correlations on the pipe interior for free/natural convection to use so I can back out an internal heat transfer coefficient for an analysis of the heat transfer under varying conditions--this will allow me to construct a revised set of guideliens for the operators. Hope to do some on-site testing to verify.

Many thanks!

 

[racookpe1978]

Good,fast answers .... Thank you!

I'm a little skeptical about convection losses though. Conduction from liquid (inside pipe) through liquid-to-wall-transition through wall through wall-to-fill-dirt-transition to final dirt temperature seems key. Seems that conduction would drive more than the convection of liquid in the pipe itself? That is, convection conditions would require a delta T to drive flow around inside the pipe .. and that delta T would have to be either bottom-of-pipe to top-of-pipe difference to drive flow in a small loop around inside the pipe, or region-of-pipe to region-of-pipe to drive flow from region to region.

"Pure" conduction would assume a stable solid mass that has no internal flow - equally unlikely in the real world, but maybe a better approximation?

How many days of topside temperature, and outside-the-pipe-at-pipe-depth temperatures do have recorded yet? Can you begin getting 12 (every two hours) hourly readings over several days when no flow in going on? Or, if you have the money and availability of people/recorders 24 hour readings for several days or weeks?

More important, do you have pressure readings of the remote end of the pipe over at least 2x 24 hour periods with no flow? If, for example, with no flow, you get equal pressure readings at 1400 each afternoon (hottest time of day), compared to consistent lower readings at 0400 (coldest time of night) then you've begun to make your case.

Things are complicated significantly if you are near-constantly flowing liquids through the pipe, since the "same volume" won';t be trapped in the pipe to cool down. (I can hear it now, "Sorry Colonel, we have to shut down all flight ops for four days to measure ground temperatures for a while .... But they could gas up at the next airfield after they takeoff if you want."

 

[JKEngineer]

Just caught up with this.

If I understand correctly, you are testing pipe with hotter than surroundings liquid over a period of time of about 1 hour.

It has been suggested that you use T probes for the fluid, which makes sense, but could be both a nightmare to implement and still leave some room for doubt, if the T varies along the length of the pipe.

Would it be possible and would it make sense to pre-condition the fuel to the ground temperature before the test? This could be done externally or by idleing the pipe for an extended period (likely not attractive). If the test is short enough, the time variation in the ground T should be less important, and you are already reducing its impact by starting with ground T fuel. You could also consider, if at all practical, shading the ground for a few feet on either side of the center line of the pipe to reduce the ground T variation. I'd suggest using a shade raised off the ground and one that is reflective, like aluminum foil.

HTH
JK

Jack M. Kleinfeld, P.E. Kleinfeld Technical Services, Inc.
Infrared Thermography, Finite Element Analysis, Process Engineering

http://www.KleinfeldTechnical.com

 

[MRMaixner]

JKEngineering--Thanks! Unfortunatley, the 1-hour test needs to be conducted inbetween flight operations, so that it wouldn't realy be feasible to let the temperature settle out to more closely approximate that of the surrounding soil. I'm thinking that by treating the fluid as a "solid" for a 1-hour test might not be too far off the mark in this application. I agree that the variation in ground temperature over the span of the test ought not be that substantial, aside from, perhaps, in the immediate proximity of the pipe.

 

[IRstuff]

I'm confused as to the purpose of the test. Why buried an uninsulated pipe when that can greatly affect the results of the test in the first place? It sounds like this is some sort of lab test, yet, it's performed in an uncontrolled environment.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[JKEngineer]

You could fill a "day" tank with fuel and adjust its temperature to approximate the ground T instead of being at 95F, which I am assuming is being driven by the outside summer temperatures in Las Vegas. Then provide that fuel to the pipe as you fill the last x number of planes, sufficient to fill the pipe with the cooled fuel. Then run the test. Alternatively, put a cooler in line with the entry to the pipe so the fuel is cooled as it is delivered to the pipe for delivery to the planes. When you stop the delivery, you are ready to go.

IRStuff: I think what has happened is that a system was put in place to deliver fuel without concern for being able to hydrotest it for leaks after initial commissioning.

JK

Jack M. Kleinfeld, P.E. Kleinfeld Technical Services, Inc.
Infrared Thermography, Finite Element Analysis, Process Engineering

http://www.KleinfeldTechnical.com

 

[MRMaixner]

Actually, this is a pretty standard test for many military and industrial applications, such as pressure testing boilers and other pressure vessels to verify integrity. It's done on a much more frequent basis on things like a fuel transfer line to verify leaks, which could have dire environmental consequences. In many of the other instances where hydrostatic tests are required, water is either the normal working fluid, or its use will not "contaminate" the system (water in any aircraft fuel could easily cause a flame-out....not good at any time during flight :-( ); this also must be performed on a short turn-around basis, since the system is used almost 24/7/365 to support USAF flight operations--no time to dump the hydrostatic test fluid, flush, test for water, dump/flush/retest, etc.

The idea of a day tank with some sort of cooling system is not a bad one.....that is, perhaps, a longer term fix. I still need to provide operators with a short term set of criteria for allowable pressure drops, albeit with relatively warm jet fuel.

 

[JKEngineer]

If a day tank or heat exchanger in line with the supply system is too long a project then I think you are back to using temperature probes. these would be best in the fuel, which is the nightmare I referred to earlier. An alternative would be to back up what temperature readings you can get with heat transfer analysis and an experimental rig.

My finite element analysis packages include computational fluid dynamics capability. Any analysis would need good values for the characteristics of the soil, pipe, and fuel.

JK

Jack M. Kleinfeld, P.E. Kleinfeld Technical Services, Inc.
Infrared Thermography, Finite Element Analysis, Process Engineering

http://www.KleinfeldTechnical.com