I'm not trying to be snarky, nasty, judgmental, etc., here, just want to convey the situation and information as I see it, being clear. The possibility of doing significant damage to a large array of expensive HDD's by cooking them bothers me.
We semi-frequently see people post about converting a 2U-12bay or 4U-24bay server chassis to FreeNAS and they want to use it at home. Whether it is because they're used to consumer PC's and have drunk the "silent PC" kool-aid or because they've seen some doofus do it on YouTube/blog/etc, quite often they want to "silence" the noisy screechy (all accurate) evil server grade fans by replacing them with wimpy "silent" fans.
Silencing a NAS is TOTALLY possible. You just need the right chassis. Or a trick strategy, But these 24-bay chassis were designed for data center use and are not a good starting point. You CAN run 12 drives, staggered, in a 24 bay rack chassis, replacing all the fans with quiet ones, and come out with a successful quiet server. The reason this works is because you have 12 open bays allowing easy airflow by the remaining drives, carrying away heat. Totally works. The moment you add the 13th drive, you get a hot spot though. You're really better off with a non-rackmount chassis that just has great airflow.
Supermicro uses some beefy 7K RPM fans, their part# is FAN-0127L4, an 80mm x 38mm thick fan, often a Nidec V80E12BHA5-57
. Three of these make up the fan bulkhead. Nidec lists these as 7.2W fans with static pressure of 270 Pa. Generating static pressure differentials tends to chew energy because of the work involved.
Now the bad news. We don't have much experience with the Norco's here. In the early days of this forum, we saw some of them come through, often with bad choices having been made. I've been pushing people towards the Supermicros because the solution was designed by a professional server design engineer and it is known to work. I can tell you that users have found the 120MM Norco fan bulkhead is the more challenging option. Their 80MM fan bulkhead is a better choice because you'll have a larger number of fans to pick from that are suitable.
Typically you would need a manometer and a full array of sacrificial drives to act as a dummy heat load/temperature sensors to measure how effectively a given combination of bulkhead/fans is actually working.
I don't have a magic suggestion for you. This is a complex multivariable problem. Cooling isn't as hard if you're using 5400/5900RPM drives instead of 10K/15KRPM drives, if you are running the server in a cool environment, etc. I'm sorry I can't give you a more directly helpful answer. Best I can do is let you know about the potential sharp edges.