#6 Fuel Oil - Thermal Expansion issues

[Bourbon103]

There is an existing relief valve (1/2" x 3/4") on an 8" NPS #6 Fuel Oil header. The header is traced with 75# steam. The header is ~545 ft long. I am sizing the relief valve for thermal expansion and getting 273 GPM required flow, per API 521-2007, pg. 33, which in my experience is WAY more flow than I'm used to seeing for a therm. exp. scenario. I have used the following numbers:

Specific heat = 0.44 BTU/lb/F
cubic expansion coeff. = 0.0004 1/F (per Perrys Handbook)
specific density = 0.96
heat transfer rate: I did a heat transfer problem through the pipe using: hwall = 2*k/(di*ln(do/di)) then heat transfer rate = hwall*A*del. T with del. T being 100F and A being the surface area of the pipe.

Is there something I'm missing/Am I being too conservative with the heat transfer? It appears that the fact that it is 4xs more expansive than water and the large piece of blocked in pipe lends its self to a large required flow.

Thanks again.

 

[Latexman]

By using just the conduction through the pipe wall, the transfer rate assumes h_inside, h_outside, and h_fouling are all infinite (1/h = 0). A blocked-in line will be close to natural convection heat transfer, which is nowhere close to infinite.

And, the steam tracing does not touch the entire area of the pipe, only a fraction of it.

I'd recommend a review of steam tracing technology to get a better estimate on heat input.

Good luck,
Latexman

 

[Bourbon103]

I agree.

I did find this which is very helpful:

http://www.cheresources.com/steam_tracing.shtml