Harvesting the CIM feed gives this exception:

org.openrdf.query.MalformedQueryException: Lexical error at line 1, column 66. Encountered: "%" (37), after : ""
        at org.openrdf.query.parser.serql.SeRQLParser.parseQuery(SeRQLParser.java:68)
        ...
Caused by: org.openrdf.query.parser.serql.ast.TokenMgrError: Lexical error at line 1, column 66. Encountered: "%" (37), after : ""
        at org.openrdf.query.parser.serql.ast.SyntaxTreeBuilderTokenManager.getNextToken(SyntaxTreeBuilderTokenManager.java:2178)
        ...

I have confirmed that this problem remains when "%" only exists in <description> elements in the CIM document. Ingest still fails for document http://q.cmip5.ceda.ac.uk/cmip5/simulation/fd70b8ea-7ae6-11e0-bb87-00163e9152a5/1/