If Dapper is binding columns by name, then the columns must all have unique names - duplicate columns be bound to the final value, with any earlier values just wasting bandwidth. This usually happens as the result of join operations, in which case you should probably alias one of the columns.

Bad:

select a.Id, a.Name, b.Name
from -- not shown

Good:

select a.Id, a.Name, b.Name as [Category]
from -- not shown