Liquid hydrogen (LH2), which has a boiling point of 20 K at atmospheric pressure, needs to be stored and transported using high levels of insulation to limit heat ingress. Typically, this is achieved by a double-walled tank with multi-layer insulation (MLI) in the interstitial space, combined with a high vacuum. If the vacuum fails from the outside, heat transfers from the air breaking the vacuum to the LH2, warming the LH2 and cooling/condensing the air. This paper describes experimental work where the vacuum of an LH2 tank was intentionally failed using nitrogen as a surrogate for air. During depressurisations of the inner tank after vacuum failure, saturation conditions were not achieved under all conditions tested, but the behaviour suggested that saturation conditions could be achieved in the vent line if the depressurisation was fast enough from a high enough pressure. This could mean that vapour may condense in the vent line for a short period of time during depressurisation. The mass flow of hydrogen out of the inner tank after vacuum failure was calculated from the data recorded, which was ~2.5 times lower than expected from the international cryogenic standards for the conditions tested. When the inner tank was left open to the atmosphere, a steady-state pressure was achieved in the outer tank, where the rate of nitrogen condensation was balanced by the nitrogen that was metered into it; this allowed the heat transfer from the inrush of nitrogen and the contribution from condensation/freezing to be calculated. The MLI sustained damage during the experiments. On examination, it was concluded that the damage was sustained during the experiment to the edges of the MLI blankets and around penetrations through the MLI required for inner vessel connections. In all locations observed, coverage of MLI was found, even if the number of layers was likely reduced in places.
Goff et al. (Thu,) studied this question.