Cryogenic Information Library
Heat Leak in Cryogenic Transfer Lines

The following information is reproduced from Dr. Thomas Flynn's "Cryogenic Engineering" 1997 Marcel Dekker, Inc., pp. 64

Heat leak in BTU / Foot
for typical rigid vacuum Insulated pipe

Pipe Size (Inner X Jacket) Inches	Fluid
	LN2	LH2
0.75* X 1.5	0.38	0.41
0.75 X 2.0	0.43	0.46
1.0 X 2.5	0.48	0.52
1.5 X 3.0	0.58	0.63
2.0 X 4.0	0.79	0.85
3.0 X 5.0	0.99	1.08
4.0 X 6.0	1.29	1.41
6.0 X 8.0	1.66	1.84

* O.D. of tube. All other sizes are nominal pipe sizes. (NPS)

In a typical vacuum jacketed pipe system the various fittings, spacers, valves, and bayonet connections will each amount to about 1.24 equivalent feet of pipe.

The above table gives the expected heat leak per foot of pipe for both liquid nitrogen and liquid hydrogen as a function of pipe size.  The flow rate is not a parameter.  This is the case because fully developed turbulent flow is assumed.  For laminar flow to exist, the pipe diameter would have to be extremely large, resulting in excessive capital cost per foot and unnecessarily high heat leak per foot.  At turbulent flow conditions, the controlling thermal resistance to heat leak is that of the vacuum space.  Solid thermal conductivity through the pipe walls, as well as the heat transfer coefficient to turbulent cryogenic liquid, are both negligible compared to that of the apparent thermal conductivity of the vacuum space.  Thus, under conditions of use, heat leak per foot of pipe is completely determined by the effectiveness of the vacuum insulation alone and hence depends primarily upon the pipe diameter.

The table shows very little increase in heat leak in changing from liquid nitrogen to liquid hydrogen.  The two heat leak conditions are within about 10% of each other even though the temperature of the inner pipe has dropped from 77 K to 20 K.  From simple relative heat transfer calculations alone we would expect the heat leak to increase by a factor of 16 = (77/20).  However, the actual case is that the hydrogen temperatures do a much better job of cryopumping than do the liquid nitrogen cooled surfaces.  Thus, at 77 K, there is still some residual gas conduction, which is not present at 20 K.  Hence, the apparent thermal conductivities are not that much different.

Back to previous page