1. μ-beam Absorber 1Absorber 2Absorber 3RF 1RF 2 Vacuum LH2 Warm window Windows in MICE: Option 1 (as currently designed) ISIS Safety officer’s comment: If absorber vacuum cryostat leaks, air is cryo-pumped and condensed on the LH 2 cold window and other cold surfaces. Then if LH 2 window leaks or breaks, hydrogen is in direct contact with oxygen. Double-fault-scenario safe concept must be implemented ! Detector 1 Detector 2 Detector gas Cold window 14 windows in total

2. μ-beam Detector 1Absorber 1Detector 2Absorber 2Absorber 3RF 1RF 2 Vacuum LH 2 He Detector gas Windows in MICE : Option 2 Is it what we need from the LH2 safety point of view? Warm window RAL’s comment: This scheme is the closest one to the design of RAL hydrogen moderators which have been approved by RAL/ISIS Safety officers. Questions remain to be answered: 1) if the LH 2 window breaks and a jet of cold hydrogen touches the intermediate safety window which is warm, how does the latter one behave? 2) Heat transfer between the intermediate and the internal windows. Cold window 20 windows in total

3. μ-beam Detector 1 Absorber 1 Detector 2 Absorber 2 Absorber 3 RF 1RF 2 Vacuum LH 2 He Detector gas Windows in MICE : Option 3.0 Warm window Coil The idea behind this approach is to separate the flip-coils cryostat vacuum and LH 2 absorber vacuum (thus providing an extra jacket for LH 2 ) and to keep LH 2 safety window relatively cold (thus eliminating question/problem from scheme 2) without increasing number of windows. Cold window 16 windows in total

4. μ-beam Detector 1 Absorber 1 Detector 2 Absorber 2 Absorber 3 RF 1RF 2 Vacuum LH 2 He Detector gas Windows in MICE : Option 3.1 Warm window Coil This a modified version 3 in which the flip coils vacuum volume is completely separated from the LH 2 volume. As a result, the LH2 is separated from the external air by 2-surface barrier everywhere. Cold window 16 windows in total

5. μ-beam Detector 1 Absorber 1 Detector 2 Absorber 2 Absorber 3 RF 1 RF 2 Vacuum LH 2 He Detector gas Windows in MICE : Option 4 Warm window In this scheme there is no path for air into LH 2 vacuum volume. But taking into account all the feed- throughs to be made, is it possible to achieve this concept in a real system? Safety window to the detector must be warm if detector is at atmosphere pressure. Safety window to the RF could be cold but dual system is not so desirable. Leakage of H 2 into RF vacuum is a question. Cold window 14 windows in total