6 replies to this topic

[Phjamo]

Hi,

We are providing approximately 4000ft of 90psi steam tracing to a tank that is 17ft long and 10 ft diameter. the entire tank is made of 1" think stainless steel. Steam tracing is 3/8" od stainless steel tubing. The tank will be insulated with 4" of (fiberglass I believe) insulation. The purpose of this exercise is to keep the inside temperature high enough to prevent condensation. the outside air temperature around the tank is minimum 20 deg F. both ends of the tank can be opened for materials to be inserted into it, so the possibility of the tank insides being exposed to 20 deg F is likely for short periods of time.

Three questions...
1) What will the temperature of the inside of the tank walls be while the doors (tank ends) are closed? (I am assuming they will be eventually the temperature of the steam 331 deg.F)
2) If the doors should remain open, how long, if ever, will it take to cool down the walls of the tank enough to allow condensation to form assuming that the 4000ft of tracing is running full bore.?
3) what is the amount of heat produced by the tracing per foot...is there a simple calculation for this?

Thanks,

Phil

[Art Montemayor]

Phil:

Yours is a problem that involves some laborious, complex, and lengthy engineering calculations - that also require extensive and detailed basic data involving the application.

We certainly can't answer your specific question without having had all the required information and done all the necessary calculations. Have you done anything on this?

One point about what you propose: installing steam tracing on a horizontal cylindrical tank means you have to install some unique tubing configuration to ensure that all steam condensate inside the tubing will drain POSITIVELY downward by gravity and into strategic steam traps. Otherwise, the system won't work. The tubing will have to be cemented to the tank shell as well, in order to ensure positive heat conductance.

[kkala]

Something is not clear to me reading the post: 4000 ft of steam tracing tube wound externally around the tank means 4000/(π*10)=127 tube rounds along a length of 17 ft, or 127/17 ~ 7.5 tube rounds (3/8") per feet; or about 25 tube rounds per m of length. So ~ 24% of tank lateral area will be covered by the tracing tubes (if my understanding is correct). This seems too conservative, even though my experience on tracing is insignificant.
It is probably better to use tracing tubes of bigger diameter (and consequently shorter overall length), which could also alleviate other problems (plugging, condensate removal, leakages). An external heating coil (steam 160 psig, 550 oF) was used here (10 years ago, it works now) for a 24 m3 tank of fuel oil to keep its temparature at 150 oC. I cannot find data sheets, but PID indicates low density (tube rounds per m of height). If density (tube rounds per m) were so high, I would have remembered the comments.

Edited by kkala, 24 February 2011 - 05:16 PM.

[Phjamo]

 end tracing.pdf   262.15KB   296 downloads  tracing REVISION 2.pdf   235.56KB   284 downloads

Please see how the tracing is designed.

Phil

[Art Montemayor]

Phil:

I don’t know how much design work you have already put into this project, but I presume you have had the assistance and consultation of a recognized heat tracing specialist firm.

I have already done what you are proposing – except my vessels were not as thick as yours, nor as large. I did both horizontal and vertical SS pressure vessels with approx. 3/8” thick walls. All my tubing was ½” OD SS and I used cement on the tubing-vessel joint.

All my tubing was designed and installed on a positive, self-draining method: there were no loops that trapped liquid condensate. The length of the loops was such that we didn’t allow the condensate to fill the tubing and cause a liquid hold up and a loss of heating due to no steam getting through. All tubing was run in such a way that we minimized the loops and maximized straight runs.

I used the following steam tracing specialist as a source of the cement and the installation instructions. You can obtain free design software from them as well as technical assistance. My design and installation worked right from the initial start-up, just as calculated - no problems.

http://www.thermon.com/US/default.aspx

[kkala]

Please see how the tracing is designed

Thanks for the information clearly showing arrangement of tracing tubes. The case of fuel oil tank mentioned in the previous post was of simpler design, a tracing tube spiral along the vertical tank of a few rounds totally. This is what was said about it. Unfortunately I have not found data sheet for more precise data.

[PreciseThermal]

Phjamo,

I do not think you need to use so much tracing. You can almost jacket the entire tank for the labor it will take to install the tracing (depending on where you are). I would recommend either full jacketing if you do not want to take the time to run a bunch of thermal calculations or leave it to tracing experts.

You may want to contact QMax Industries, Inc. ( www.qmaxindustries.com ). They can run calculations to model steam tracing and typically reduce the amount of tracing required. It would be worth calling and seeing if they can help.