You don't seem to understand a few fundamental things.
You have NOT defined the specification, material, nor torque being used to install these bolts.
"Ordinary bolts" is NOT describing ANY design feature of the bolts!
Go measure them again: They do not scale at M24 based on the size of the text and embedded/stamped label information for the manufacturer in Indiana. Yes, they may be metric - even though made in the USA. But these bolts do not appear to be big enough to be M24.
Your bolts are breaking because the initial torque applied during assembly, and the added load imposed on them during operation - increased by unknowns you refuse to tell us about such temperature, environment, chemistry, start-stop cycles or sudden dynamic loads! - are greater than the ability of the installed bolts to withstand. It is as simple as that.
To continue a "lessons learned" criticism: Yes, lubrication during assembly is ESSENTIAL. However, the wrong lubrication in the wrong environment applied to the wrong alloy bolt-nut-base metal combination with the system fluid or chemistry "MAY" caused further harm. Or it may not. We do not know enough to judge at this point. Your photo shows no sign of lubrication, of torque applied, of turn-of-nut-records, or of ANY indication that a sequential torque is being applied around the joint.
At that size of the claimed large M24 bolt and nut, the METHOD of torquing is critical: You should be applying the torque (which is needed to get the pre-load required for successful operation) in 3 or four stages towards maximum torque (of 25%, 50% 75% 90% and final torque) by a qualified operator using a calibrated torque wrench in a star or hex pattern around the entire joint. Is this being done properly? Have you inspected the operators ACTUALLY DOING the installation? If they are using an air impact wrench to "just drive them down dry" - then, YES. THE BOLTS WILL FAIL.