In ADO.NET, correctly handling null is a constant source of confusion. The key point in dapper is that you don't have to; it deals with it all internally.

• parameter values that are null are correctly sent as DBNull.Value
• values read that are null are presented as null, or (in the case of mapping to a known type) simply ignored (leaving their type-based default)

It just works:

string name = null;
int id = 123;
connection.Execute("update Customer set Name=@name where Id=@id",
    new {id, name});