Do you have the bootable OS/2 disks for the system? Those may either be necessary (or at least helpful).
Have you checked the hardware itself for issues (failing drive, overheat, possible BIOS setting change when the machine went off)? I've seen (at least a few times), where a machine running for as long loses power, suddenly won't reboot, and it's simply because something in the BIOS reverted to some default state due to a failed CMOS battery (which is quite possible after 8+ years).
So, you may wish to do the following (see note below)...
[1] Check the BIOS for any changes (ensure that any memory option is NOT set to OS/2)
[2] Try a "Live CD" (eCS demo, BartPE, Hawke PE, a Linux Live CD) to see if it will boot on the hardware
[3] Find a diagnostic disk to test the hardware (Microscope 2000, IBM/Hitachi Drive Fitness Test, etc)
[4] Boot off the OS/2 floppies (Warp 4 and earlier)/Boot CD (Warp Server for e-Business and eCS) and check out the partition for any errors... (it is possible something got damaged during a checkdisk when the machine got restarted.
And possibly before any of that, you may wish to install a secondary drive (need something smaller if this is Warp 3 or Warp 4 without the latest fixpack), then boot off the OS/2 disks/CD, go to commandline, create a partition on the second HDD, format it (HPFS), and back up the original drive's contents to the new drive (then shut down, and remove the new drive).
Doing so is simple. Let's say that FDISK and Format assign the new drive the letter F:
At the command prompt, change to F: (ie: F: <ENTER>)
Type (assumes the original Drive is C:) xcopy c:\* /h /o /t /s /e /r
Wait till it is finished (note if it shows any errors - they will probably be read errors if it does, which would be indicative of a failing drive/media - if that happens, manually xcopy every directory you can...
xcopy c:\SomeDirectory\* /h /o /t /s /e /r
xcopy c:\SomeOtherDirectory\* /h /o /t /s /e /r
etc.
To the xCopy command, you can also add the switch /v if you want it to verify the files on the new drive as it writes them (slower, but a good idea if you suspect some component in the system may be failing).
If you suspect it may be a hard drive, then doing the backup first may be of paramount importance. I have seen various dying hard drives that barely last through the "rigorous" access of doing a backup, and quite a few that have totally failed during hard drive diagnostics.
And of course, when you run any drive diagnostics on the drive, ensure you do not pick any test that is destructive (many write tests may be... and some write tests that shouldn't be, can still end up being if it cannot successfully write the data to it's new planned location).
In addition, write tests with some HDD diag tools on HPFS is not always a good idea.
Sorry so verbose... but without more of an idea of what it is hanging on, it's hard to narrow down what advice to post...
Rob