This year we’ve asked our intern teams to share their perspectives on their work at BPL on our blog. Look for monthly updates from the teams – beginning with this post from Michele Bradley, Ingrid Thvedt, and Bahti Zakirov of the DCAS Energy Data Lab team.

Reducing carbon emissions by 80% may seem daunting, but it’s what the city of New York has decided to do. The 80 by 50 plan is a goal to reduce greenhouse gases 80% from 2005 emission levels by the year 2050. As part of this plan, Mayor de Blasio has established a ‘Carbon Challenge’ to New York businesses to reduce their carbon emissions by 30% or more over the next 10 years.

Buildings are a huge part of New York’s city energy use, making up 75% of the energy use of the entire city. Focusing on buildings instead of other areas, such as transportation, can make the biggest impact overall.

Here at BPL, we work towards making this possible. By modeling the energy use of buildings, we

  1. edlBenchmark current energy use of large buildings: Before improvements can be made, one has to know the current energy and fuel use. Finding this baseline is the first step to sustainability. As shown in the graph at right, we can find this building’s baseline usage with Change Point Linear Regression Model using Inverse Modeling. It is the amount of usage a building uses irrespective of temperature.
  2. Compare the relative performance of buildings: By looking at the performance of different buildings, we can see which ones have the most room for improvement. Comparing similar buildings can also be inspiring– if one office building can be energy-efficient, then so can others!
  3. Discover areas for improvement: Modeling energy use based on temperature allows us to estimate the efficiency of the heating and cooling systems. If we see any atypical behavior, such as the overlap of heating and cooling, we can make suggestions on how to fix it.

In researching how buildings use energy, we are integral to achieving this goal in New York. The Carbon Challenge is just the beginning.

Share →