Some EIMs are mutually exclusive or mutually dependent. See the EIM section: Packaging Measures for more information.

Some EIMs are mutually exclusive or mutually dependent. See the EIM section: Packaging Measures for more information.

We use multiple proprietary sources to access the county’s stored information for each building. See the Building Stock and Portfolio Studies section: Insights with just an Address for more information.

Our publicly accessible primary data sources - the DoE / NREL / PNNL Commercial Reference Buildings and Residential Prototype Buildings have publicly accessible accuracy studies that show that they line up with the aggregated performance of the buildings they represent. We make modifications to this dataset to apply them at the single-building scale (as opposed to the aggregated scale).

In accuracy studies we’ve performed, the average accuracy over publicly available datasets is 1%-2% error, showing that these models, too, perform very well when representing the average building of its class.

For example, our modelled building EUI for a Medium Office in ASHRAE Climate Zone 5B, built in the range 1980-2004, with Natural Gas as the primary heating fuel will fall within 1%-2% error margin of the average of all the buildings in the U.S. that fall into that category. To the extent that this specific building differs from the average of all buildings in that category, the error margin will increase. That difference from the average can be accounted for by taking advantage of our Calibration and Refinement features.