Just some perspective on noises. When I was a Reactor Operator at Wolf Creek Nuclear Generating station some years back we were restoring from a refueling outage. During plant warm-up there were some personnel inside Containment doing clean-up when they heard a loud BANG. It was loud enough to a single trigger seismic detection alarm in the Control Room. The heat-up was suspended and an investigation immediately started. After 3 days of inspections, nothing out of the ordinary was found, so heat-up was again started. Reaching operating temps with no other problems plant startup was commenced. We ran the whole fuel cycle with no problems. After the next refueling outage they put monitoring equipment inside Containment to see if the problem resurfaced. Sure enough we got another BANG! Rinse and repeat. The next refueling the NRC was there to watch the heatup and startup, but by that time we had instrumented the crap out of everything inside Containment so when heat-up was started if it went bang we were going to find it. It did, and we discovered that during our heat-up the reactor coolant piping had contracted on cooldown enough to pull away from the pipe shims. During heatup the piping had to grow but was being confined by the shims and when the pipe got hot enough would slip past the shim suddenly and cause the BANG. The piping in question was 29 inch ID stainless steel pipes with 4 inch thick walls, which were full of water, so pretty heavy stuff and when it moved a quarter inch in a split second made a heckuva racket. We ended up replacing the shims the following outage (with teflon coated units) and no more problemo. Mechaincally it sounded nasty, but materially had very little effect. It cost the company about $1 million to find the source of the BANG! As far as they and the NRC were concerned, it was money well spent, as far as actual plant operation and operational longevity, it was money down a rabbit hole, no effect on the plant.
During plant Engineered Safety Features testing following a refueling outage we performed a Safety Injection test to ensure all equipmemt started as designed in case of an emergency. The test went as designed but during the test a loud BANG was heard inside Containment. That bang also set off several seasmic alarms but this time it was actual movement alarms and it was felt in Containment, the Auxiliary Building, and the Communicatons Corridor. The test was started by simultaneously providing a safety injection signal and ESF emergency bus under-voltage signal. This was meant to ensure that if a SI occured and the plant lost power the emergency diesel would start and load emergency equipment automatically to maintain core cooling. Turned out when the test was performed the Plant Service Water system was divorced from the Essential Service Water system (by design) and when the safety related busses were deenergized for the test and the loads sequenced on after emergency diesel generator startup and energization of the electrical bus, enough time had elapsed that a water column seperation occurred in the drain piping from the Contaiment Coolers to the basement of the Auxiliary Building (about 13 seconds). The column seperation estimated at about 60 feet in height in a 30 inch header was collapsed when the ESW Pumps started and started flowing. The collapse of the water column (one water column slamming into a stagnent column) caused the piping to move dramatically in three buildings. That BANG was very significant and cost about $65 million to fix. They had to re-engineer the return piping to the Ultimate Heat Sink to ensure the piping stayed full when water flow was lost for a few seconds (the time between loss of AC power to the bus and EDG startup (up to 12 seconds) and re-energization of the bus plus the 20 to 25 seconds from that point to start of the ESW pumps).
So noises can be nothing or really really something. If it doesn't cause the machine to lurch, it's probably nothing.
