Assigning an emissions impact to changes in electricity demand will be ever more important as we value carbon appropriately, but this value depends on the method used. How well do these methods fair against a power grid model?