The actual number seems to be on the order of ~2*75kW for conventional AIP (Gotland) and ~1.2MW for nuclear AIP. This is the number the original Twitter poster got assuming a 10 MW thermal reactor driving 4*320kW Stirling.
If this is truly a mini-nuke, it would be a strategic platform. I doubt they'd give us something as classified as engine power before first boat even gets launched. 10MWt delivery 1.2MWe would be 12% efficiency. That is stunningly low. Even 091's reactor going directly into shaft achieved like 18% efficiency.
Yes but the type and size of rafting will be different. It all depends on the specific vibrational spectrum.
Atmospheric management is doable. It's just electrolysis of water to produce oxygen for breathing purposes. You'll need to produce 15 L per hour of oxygen per person.
A 1 MW electrolyzer produces 50 L/min or 3000 L/hr which is enough for a crew of 200.
For a crew of 36 as typical on a 039A SSK, you would only need 540 L/hr of oxygen production which scales down to 133 kW. If that number is further cut, then you can get away with an even smaller electrolyzer.
Again, it's doable, but we have to think about what is the requirement here. Different requirement will lead to different sized propulsion unit.
Let's do some thought experiments here
If PLAN wants this to have real endurance/stealth, then it needs to have atmospheric control machinery, which takes power to operate. 300kw is required to keep lights on for Yuan sub and maybe move occasionally in deep waters. If we add more powerful equipment like atmospheric control, more computation power & sensors, the sub will also be a little larger. Maybe 600kW is needed to keep the lights on at low depth with everything you need and not moving.
Let's say for a 4000t sub, you need 25MWe to move at 30 knots under water (at reasonable depth) (basing this on VA needing 30+MWe to achieve some classified speed). I could be over or under-estimating things here.
To sustain 10 knots, you'd about 1MWe to do so. speed is correlated to power ^ 3
So, you probably need 1.6MWe. Now if we consider eMotor/Drive to be 80% efficient, then we need about 2MWe generated by steam turbine/Stirling to achieve at best 10 knots speed. This is actually better than I thought.
But it rests on my assumptions about speed/power being correct and keeping the lights on calculation to be correct. I could be entirely too optimistic about both here.