Either in the printer > Adobe PDF > properties > settings tab, select the PDF/A settings file (don't remember the name, but it should be obvious). If you are using PDF Maker in WORD or such, select the create PDF preferences and again select the PDF/A settings file.

If you have created a PDF, you can also use preflight and select the convert to PDF/A script.

(I need to go to my other computer if you need names of scripts and such.)

To help clarify the OCR a bit. You can only OCR an imaged PDF one time. If the PDF is not a set of pure images it can not be OCR'd. OCR stands for optical character recognition and once it has been done there is no need to redo it. There are multiple options with the OCR. 1. Clearscan replaces text it finds with the best guess. If you were wanting to convert to a text document and deal with editing, this would be the choice. 2-3. Searchable text leave the image alone (maybe what you want to keep the original time sheet view intact) and adds searchable text in a layer behind the image. This text allow the search capability.

That being said, there is not real sense to retain OCR capability if you have OCR'd a document. The only question might be if the search capability would be retained. As I mentioned before, you would probably use the preflight to do this conversion to PDF/A.

I just did an OCR (searchable on a graphic PDF). I then went to preflight and use the conversion to PDF/A. I first tried the 1a version and got errors and then repeated with the 1b and it said ti worked. After that the file was still searchable. I used AA9 for reference. With that said, it would seem that the process works as I understand your need.