Are you sure the contacts are actually sticking, i.e. do they stay stuck even when not powered? Reed switches are operated magnetically. Even though you are in a still tube, level fluctuations could be "chattering" the switch and causing the iron strip inside to magnetize semi-permanently. The fact that they felt it necessary to debounce the circuit with a 5 minute timer makes me think that this was a problem they were trying to solve. The problem with that solution is that it only solves the control issue, the reed switches themselves may still be operating too frequently. Here is a web site that describes how reed switches work. I have found that as reliable as they are, they cannot be used for high speed high switching count operations because of that magnetic memory issue.
http://www.osdc.co.jp/english/products/products.html Or is it just that the timers are not de-energizing when they should? I have definitely had 24V circuits get enough energy induced on them to keep solid state circuits energized, even if it wasn't enough to energize them in the first place, essentially a seal-in of the circuit. A small resistor across the terminals of the time coil should take care of that.
The simple test would be to see if everything goes back to normal as soon as you take power off the circuit. If it is an induced voltage issue that is keeping the Idec TDRs from dropping out, they should drop out again as soon as power is removed. If the reed switches are sticking closed, they would stay that way until the residual magnetism dissipates.
