Because allowable stress of fracture sensitive steel materials are coincidentally governed by fraction of tensile (1/3, 1/3.5) rather than 2/3 yield, and yield is typically ca 55% of tensile, hydrotest primary stresses need not normally be near 90% of yield. The recommended safe temperature for resistance to brittle fracture, which goes with the lower expected hydrotest stress level, is 30°F above the MDMT for material curve and thickness. I would suggest considering whether this is adequate if going to 90%. When going to 90% of yield specifically with a fracture sensitive steel, you might be way out of bounds in fracture safety. If you do the fracture mechanics for plausible flaw sizes, you might find the material is incapable of the required toughness, unless very high temperatures relative to the curve are used. The 90% provision was first used with Div 1 vessels to allow large field fabricated low pressure pipe stills and fractionators to be tested, a once in a lifetime event. It was a big deal event with a lot of sctrutiny and great observation distance during the test, I assure you. I would be very cautious about pressure testing just any erected vessel maybe 25 to 50 years old at the same 90% stress level. Div 2 more specifically governs maximum test stress but has a whole range of supplimentary reinforcement and other calculations, and stress limitations for other than primary stress, that goes with 90% allowance. Just allowing 90% for primary stress is an abridged story. If you do it, consider doing it to Div 2 even if a div 1 vessel and consider defects like fatigue cracks and corrosion that may have occurred in an old existing vessel.