Win 8.1 File Content Indexing

I have an issue at a client that has thousands of invoices stored in folders, in PDF format. The often need to look up invoices based on invoice numbers, or other details (such as vehicle registration number) etc.

I have added these folders as a location to be indexed, installed the latest Adobe Reader, and made sure the ifilter value in the registry is the correct one, and confirmed this is available under file types to index. PDFs are set to have content indexed.

PDFs are created by printing to the Bullzip printer driver.

This results in a PDF with textual content, with a few blocks of text (like an invoice) holding customer details, job details, and contact details.

Now, the problem arises when they search, in that when the search for an alphanumeric string, such as a vehicle registration number, results seem to contain strings found as if different boxes of text are being joined. E.g. part of a registration number, and part of a number from another area in the file, being combined, and then matching the search criteria.

Could this be caused by something wrong in the way search is setup? Or in the way the files are created?

This all worked perfectly when they were running it all on Windows XP