It could be a couple of things, either something changed at the host file system layer, or possibly free space in the DDS being slow to reclaim due to the firmware level and older hardware.
These appear to the be the offending volumes and are all dedupe volumes:
Vol1_PRD_BDKC1
dboracle1
dboracle2
Vol1_PRD_BDIC1
However since both reserved and used space are pretty close it could be that some large changes have occurred at the file layer and the array hasn't been made aware of these. First thing I'd try would be to run an unmap to signal the reserved space can be reclaimed.
Have a look at this WP appendix A for reclaim techniques :
https://h20195.www2.hpe.com/V2/Getdocument.aspx?docname=a00006358enw