Conclusions

\label{conclusions}
With LL84 and PLUTO data a robust model was built to predict the EUI of multifamily and office buildings in NYC. This model was built as means of defining a market specific performance metric for energy efficiency for NYC buildings. This score will complement the EPA’s Energy Star scoring methodology which is based on national level benchmark. The developed model does not explain completely the building energy use intensity and neither was that the goal of the model. The objective of this model was to provide a city-specific baseline for the prediction of EUI to be used as reference to identify the better and the worse performance buildings in the city. In that sense it was necessary that some of the variability was left unexplained by the model. Afterall, a major part of buildings’ energy efficiency is dependent on human behavior for which there is no explanatory feature in any data set.
The scoring methodology and the fact that the MOS wants to better inform the building owners is an attempt to unlock the energy saving potentials of behavioral change. The Energy Snapshot is only a first step, but goes in the right direction of trying to engaging building owners and managers not only to invest in more efficient technologies but also to promote behavioral change.
However, in order to actually engage the building owners it is important to convince that the metric by which they are being evaluated is a fair and sensible one. In this context, the separation of the buildings into peer groups is another important contribution of this project since it addresses the mentality that some building owners might have that their building can not be fairly compared to other buildings in terms of energy efficiency because of its unique specific features. Since the buildings are being compared essentially only with similar buildings, it might be easier to convince them that the metric and the comparison are fair.