Those signals refer to the correction sent to the TEC to cool down either the pump diod and the crystals of the slave laser.

the correction moves between 0 and 100.

100 means the correction is null and "0" means that the sent correction is maximal.
In the plot we got ~10 which is nearly close to the maximum of correction.

Such high vale may increase the ageing the TEC at some point up to some breakage 

In the past the correction used to be around "50" instead (see attached doc, p2, data for a Xstal temperature setting of 25 °C gives a correction @ OUT from PID of ~50 %)

This would mean that the cooling is by far less efficient. It could be that the water flow is reduced due to the dirts seen in the water cooling system of PSL.