PROBLEM: I have been struggling with this error now for days now and have tried everything I know. I have an SQL statement that pulls data from several tables into another table. The field in question is a NTEXT field from a SQL 2000 database, which I now import into a SQL 2008 R2 table that is NVARCHAR(MAX) data type because I though the issue was the NTEXT data type. However the SSIS package that is just an OLE DB Source (with 1 field) into an Excel Destination is still giving me the Unicode and Non-Unicode Error!! Several rows of data are over 8000 characters in length.
ANSWER: After a lot of pain I finally came to the conclusion that Exporting to EXCEL is not possible so I turned to CSV. I used "Flat File Destination" object, pointed to a CSV that I had created with just the Headers. The Text Qualifier was set to double quotes. In the Columns section I set the Row delimiter to {CR}{LF} and the Column delimiter to Comma{,} because it is a CSV! The final part of the puzzle was to remove and double quotes, Carriage Returns and Line Feeds. I also had to convert the NTEXT field to VARCHAR(MAX) because REPLACE will not work with NTEXT. This is what I ended up with for the columns that had these "invalid characters".
REPLACE(REPLACE(REPLACE(CONVERT(VARCHAR(MAX),[MyNTEXTColumn]), CHAR(13),' '), '"', ''), CHAR(10),'') AS 'Corrected Output'
I replaced {CR} CHAR(13) with a space to that we could have it formatted well for the consumer. I hope this helps someone out one day.