This is a great discussion we are having on future Chinese carrier fleet size. Btw, I don't think we can use annual operating cost of an USN carrier to infer the cost for Type 003 and its air wing. PLAN crew members are just going to be paid a lot less. PLAN activities will also be less demanding than what's expected out of a USN group. Once China really gets its carrier construction going, I think $4.5 billion might also be an overestimate to production cost.
Having said that, carrier groups obvious cost a lot of money. In the past, people have often suggested that PLAN will eventually settle in with 6 carrier groups, which would work out to be 2 groups per fleet. I'm not sure there is any evidence that's how many groups PLAN planners are looking at. We have seen them having great ambitions than even the most optimistic projections. Their carrier program is progressing at much faster pace than what I would've expected. Their LHD program is also progressing faster than I would've expected. If they do end up with 10 carrier groups, I would not be shocked. It does seem to me that 6 is a more likely number until they can match up more evenly with USN air wing and nuclear submarine fleet. It would take them launching a super carrier every 4 years to just get to 6 carriers in service by 2035(that's assuming CV-16 doesn't retire by then). If we assume that can be reduced to one every 3 years after the first couple of unit, they would not get to 6 CATOBAR capable carriers until 2040 and 8 by 2045.
If USN, with 11 carriers, thinks that it only needs 273 F-35C and a larger number of super hornets, I think it'd be unrealistic to expect more J-35s than that. So, I'd say a production run of 270 J-35s is the most likely we will see. The remaining of the fixed air wing will be KJ-600s, J-15B and J-15BDs (maybe 160 of them in total).
Now, 270 is still a pretty small number for a 5th generation project. The cost per aircraft will be quite high if they end up producing 20 J-35s a year in peak production. In some years, they wouldn't even need to produce J-35s if a new carrier is still waiting to be launched. J-20 would be cheaper than J-35 to produce, operate and support in that scenario. In order to keep SAC and the J-35 suppliers healthy, they absolutely need a land version of this aircraft. They cannot make the same mistake that USAF did in shutting down F-22 production so early. The question is what would be a good number of land based version of FC-31 to purchase. I originally anticipated more land based version of FC-31 than J-20. However, given the headstart that J-20 has, that seems to be unlikely. If CAC can actually keep 4 assembly line going, 50 aircraft a year should be easily achievable. With that level of production, they will probably have 300 to 400 J-20s in service by the time they are ready to mass produce land based FC-31. PLAAF is showing an urgency for J-20s. They are not waiting along for WS-15. And even when we look at J-10s vs flankers, the annual production rate between the two weren't that far apart. I don't buy the theory that PLAAF will necessarily purchase more of the smaller aircraft. I think we are more likely to see CAC and SAC eventually building about the same number of 5th generation aircraft per year. If we make the assumption that they will mass produce FC-31 line of aircraft for 20 years (2025 to 2045), that could be up to 800 over its life time, or about 500 in PLAAF. You might have 800 J-20s built from 2020 to 2040. That would be a pretty large run of 5th generation aircraft. I think they have the budget to do so. During peak years, they might procure more 5th generation a year than America does.
