I don't know much about the AIC ratings on OCPDs except that it has to be equal to or greater than the fault current available from the utility.
So, when a residential main has 22k AIC does this mean that each leg of the service is potentially capable of sourcing 22,000 amps? I would assume so, because all the single-pole breakers must have a corrosponding rating. Does this mean that the actual fault current between legs could be as high as 44,000 amps?
Also, other than the different markings, there isn't an obvious difference between 10k and 22k AIC breakers. Are the ones with the larger AIC rating simply more robust? Are they built out of stronger materials? Will they really withstand a whopping 22,000 amp fault before they fail?
The proccess around here, (I say that because it AIC is often ignored in some locals) is that you get an AIC letter from the utility stating what the Available Fault Current is at the drop. Usually there is no problems, but on occassion you can get one that is rather high, and excceeds the ratings of the breakers. 10K is the limit on resi here, and some commercial too. From there you have a few options on equipment, series rated breakers to get below 10K, or current limiting fuses. Both require a graph from the manufacturer or engineer stating available current and final AIC.
How they are built - no idea.
Mark Heller "Well - I oughta....." -Jackie Gleason
If the calculated fault current is 22,000A, it is typically a three phase bolted fault value. A breaker listed as 22kAIC is theoretically more robust than one listed at 10kAIC. Sometimes it is a marketing and liability issue though. I've heard of breakers that are manufactured with a 22kAIC rating, then relabeled also for sale as 10kAIC, so the manufacturer has one manufacturing process and pays a lower liability insurance on the 10kAIC breaker. It works out cost effective for the manufacturer.
AIC ratings of CB's are tested by mounting them inside a panel with a bolted short accross the CB output. The input current is regulated (10K, 14K, 22K, etc) and they either interupt the fault current without being damaged, or they don't (and typically blow up). If they pass, fine. If not, the CB line looses it's rating and the mfgr needs to make some manufacturing mods.
Available fault current from the poco is simply the max current their equipment will put out with a bolted fault at the lugs on the poco secondary. Some utilities have fused outputs (like So Cal Edison) and their fault currents are relatively lower (48KA max on 480V systems, 10KA on resi). Some don't (like LA-DWP), and their fault currents can be guite high. We were on a project at Ports-O-Call in San Pedro and LA-DWP gave us a fault current of something like 140,000 amps at their transformer. Seriously. Our service was about 300' away, so the wire resistance knocked that down quite a bit - meaning if a fault occured at our service, the actual fault current would be much lower than if the fault occured at the transformer.
Fualt current is different at different points in your electrical system, primarily because there is differing amounts of impedance between the source transformer and different points within the system. The specific fault current at any point is a combination of what's actually available from the poco equipment and the cumulative impedance between the poco equipment and the point you're trying to calculate for. Frankly, I just used wire resistance to come up with rough fault current numbers, the actuals using impedance would be lower.
There are 10 types of people. Those who know binary, and those who don't.
The testing of circuit breakers is actually much more complicated. They are tested in "real life" situations so they have a specific conductor attached to their load terminals. Rarely do the "blow up" when they fail the test. To pass the test they must be able to be reset and clear a second fault at the same level, the load conductor must not pull loose nor be damaged, and the gasses vented by the interruption must be controlled and not cause additional faults nor fires.
In this area one POCO always gives fault current values at their meter socket. A different POCO gives it at the terminals of their padmount transformers for underground feed and at the "drop connection" for over head services. The point is always ask.
During fault conditions motors also contribute current, so calculations become slightly more complicated in facilites where motors make up a large part of the load.