Hi Guohui,
It certainly should not be that hard. The system administrator should be able to set up the X3DNA environment variable and add $X3DNA/bin directory to command line search path, so the 'apache' process, or all users, can access them.
This is clearly an IT issue where Wilma should be able to play a crucial here. In the lab I am working, there is a (part-time) system-admin to solve such issues.
While the problem persists, you could focus on the rebuilding and analysis parts to get them done in a more decent form. As for 'blocview', the system needs to have MolScript, Raster3D, and ImageMagick installed, and properly configured for system-wide access. You could (through Wilma, perhaps) check with the NDB/PDB people -- they use 'blocview' for all the nucleic-acid containing structures.
Have a good weekend.
Xiangjun