When a model contains non-ascii characters, the eplustbl.html file in the reports directory scrambles unicode characters. Attached is an input osm file and the respective html files from the run and reports directory. You can see the unicode issues in the zone name. I believe the problem is the ISO-8859-1 encoding used here
unicode_example.zip