First, take a look at this thread over at arfcom: Bye bye EBR (warning, lots of big pics)
Taken a look yet?
Though I don't have complete information, this looks to me like a textbook case of an out of battery ignition; one of the worst failures that can occur in a firearm.
An out of battery ignition is when a round is ignited before the weapon is locked and ready to fire. Without the full support of a locked breech and chamber walls, the burning powder expands far beyond the pressure limits of the (usually brass)cartridge case, causing the case to burst. This results in a huge pressure spike in the receiver of the weapon (far higher pressure over a much shorter period of time than in a normal ignition), and against the breechface (in this case the bolt face of the AR).
Generally speaking, when a weapon fires out of battery, that weapon is destroyed; and that is clearly the case here. Thankfully no-one was injured beyond some bruising and being shaken up.
The original poster and several of the commenters are blaming the ammunition, saying it was likely a double charge, or a squib load; but I think it's clear that they are incorrect.
I’ll tell you right now, this failure most likely wasn’t the ammos fault... or at least not completely.
This almost certainly was not a double charge. You can’t charge a .223 with enough powder to cause a failure like that from an overcharge using rifle powder. Rifle powders are fairly bulky (even very fast ones), and the .223 case has a maximum capacity (with a fully compressed charge) of somewhere around 30gr or so of rifle powder; which is only a few grains more than a standard load.
It’s POSSIBLE that there was a squib load which lodged in the throat of the bore, and caused overpressure when the round behind it ignited; but in an AR platform rifle that is nearly impossible. Because of the gas system architecture a squib round would have to travel far enough down the barrel to pass the gas port, and then get stuck. If the following round were then ignited, the resulting overpressure would bulge or split the barrel, and likely damage the bolt.
From the character of the damage I'd say it's clear this did not happen; and I think it’s far more likely that this was a severe out of battery fire, and a weak, poorly made upper.
I’ve examined these pictures pretty closely; and I’m very familiar with the failure modes of the AR platform and it’s ammunition; and here's why I believe this was an out of battery ignition:
You can clearly see the initiation point of the failure is in the barrel extension (where the bolt locks into the chamber). If the failure had been caused by a squib, the rupture would have begun in front of the chamber, and most likely would have been STOPPED by the extension. Then the rear of the receiver would have been damaged by the bolt.
In this case, it’s clear the failure initiated at the barrel extension. This indicates either a massive overcharge, or an out of battery ignition.
The only way you can overcharge a .223 case that much is by using pistol powder; because there isn't enough room in the .223 case for that much rifle powder.
While it is conceivable that a factory reman (ultramax are factory remans) was loaded with pistol powder, it is incredibly unlikely. There was most likely no pistol powder anywhere near the production line this ammo was made on (for safety reasons you dont keep different powders near each other or near the wrong production line. In fact there are ATF regulations against it).
Slower rifle powders will fill a .223 case completely without excessive pressures. If there was a moderate overcharge, as would be the case using a very fast rifle powder and filling the .223 case completely (to a compressed load state); even presuming that round made it through the quality assurance inspection of the line (which is unlikely but possible), the result would have been a case head separation and some damage to the bolt; and most likely the magazine being blown out of the rifle.
In this case, the damage was far more severe and far more radical. There was a complete failure of the upper receiver, with primary failure initiation at the barrel extension radiating though the ejection port; and secondary failure induced by the bolt carrier.
Looking at the pictures of the bolt retail, you can see that the bolt and carrier were thrown backwards violently, however the bolt head isn’t sheared off, nor is the barrel extension. This can only indicate that the bolt was unlocked at the moment of failure.
This incident was clearly an out of battery ignition.
Now, as to cause, that would be more speculative. It is possible that soft primers are the culprit… and in fact I would guess that they are very likely a contributing factor. Most likely however is that the home AR builder used a, heavier, older style firing pin; and when the round had some trouble chambering (either a tight chamber, a dirty chamber, or slightly out of spec ammo) instead of simply failing to chamber and fire (as is the appropriate failure mode), the firing pin traveled forward with enough energy (or was jammed forward by a piece of grit, or by mechanical damage) to ignite the primer (either with or without the assistance of the hammer); and an out of battery ignition occurred.
This is unfortunately not an uncommon failure mode for semi-automatic rifles in general; it's called a slam fire, and it has in the past been an issue with the AR family (though most often with full auto guns), the AK family, and most machine guns (if allowed to wear too greatly, heat up too much in firing, or if not cleaned properly).
In theory, the design of the AR family should prevent slamfires from occurring. The firing pin should not be able to protrude far enough forward of the bolt face (even with the pin at full forward travel) to ignite a round, until the bolt head is at least very close to full lockup. This should both prevent slamfires, and prevent the weapon from firing out of battery.
Unfortunately this is not always the case. All manufactured goods in this world have tolerances, and there is a point at which the intersection of tolerances can cause a failure. The firing pin could be right at the long end of tolerance. The bolt and carrier right at the short end.The cam slot right at the long end. The bolt lugs right at the narrow end. The primers just at the soft end etc... Any two of these factors combined could cause this type of failure; any three would make it likely.
You might say "Oh but thats so unlikely, its a one in a million chance that those tolerances would line up like that"... well, it may be a one in a million chance on a top quality piece with tighter quality control; but not all weapons are made to very high standards... even if they were however, one in a million isn't so much when you consider the fact that between 200,000 and 300,000 AR type rifles are made worldwide every year.
Clearly, the design of the AR, while resistant to slamfires, is not immune as thousands of ARs have slamfired over the years; and at least hundreds have suffered an out of battery ignition.
I would wager that if the builder ever checked a round cleared from the chamber without firing, he would note a slight dimpling of the primer. This dimpling results from the firing pin lightly impacting the primer of the round as it is chambered. This is common to ARs with the heavier firing pin, using commercial grade primers. The problem is, on occasion the primer may be just a bit more sensitive than normal, or just a bit softer, or the weapon may return the bolt with just a bit more force... and then you get a slamfire.
This tendency to slamfire is why the military, and better manufacturers, switched to a lighter firing pin design; and also why they now use MUCH harder primers on military 5.56 ammunition.
I personally use CCI benchrest primers, or CCI 5.56 nato mil-spec primers because they are harder and less sensitive than other commercial primers; specifically to address the slamfire issue.
As to preventing out of battery ignition, my only suggestion is to use quality components, manufactured well within tolerances; and maintain them properly; and the design of the weapon itself should prevent the problem.