Reaching 80 x 50 in New York City

5 min readOct 19, 2020


When US Senator Jim Inhofe tossed a snowball on the Senate floor to prove to the American people that the planet can not possibly be warming, he represented a great grip of climate denial in American politics. While we may laugh at the absurdity, the threats of global climate change remain in the eyes of many, a topic for debate.

New York City is one of many U.S. cities that has taken on the issue of global climate change. Former New York City Mayor Bloomberg recently lauded his own record on climate during his 2020 Presidential campaign, pointing to the drastic reduction of the City’s carbon footprint during his administration, when compared to the rest of the country. While true, much of the credit Bloomberg claimed was due to the switch of power plant fuel sources from coal to natural gas, which is a less carbon intensive energy source.

NYC’s Greener, Greater Buildings Plan

According to New York City’s Greenhouse Gas Inventory, stationary energy (buildings) has historically, and continues to be, the largest contributor to greenhouse gas emissions in New York City. In 2019, stationary energy contributed 37.32 million metric tons of CO2 into the atmosphere. While down from 2018, this 37.32 million compares to just 15.63 million metric tons from transportation.

The Greener, Greater Buildings Plan (GGBP), part of the deBlasio administration’s PlaNYC, aims to further curb the City’s building greenhouse gas (GHG) emissions. Key local laws include:

  • Local Law 84 (LL84), which requires private buildings over 25,000 ft2 (modified by Local Law 133 from 50,000 ft2) and public sector buildings over 10,000 ft2 to report their energy and water consumption each year.
  • Local Law 87 (LL87), which mandates that buildings over 50,000 gross square feet undergo periodic energy audit and retro-commissioning measures.
  • Local Law 97 (LL 97), which places carbon caps on buildings, is a commitment to achieving an overall reduction in greenhouse gas emissions of 80 percent by 2050 compared to 2005 levels.

Data & Methodology

My analysis takes a closer look at the buildings subject to benchmarking under LL84 in 2019. I’m particularly interested in what the distribution of properties might suggest about the efficacy of the City’s current emissions reduction plans. That is, is the City’s emissions reductions plan effectively targeting the properties that are least energy efficient?

After cleaning BBLs and some important energy metrics in the LL84 data, I also pulled in the City’s land use and geographic data at the tax lot level from PLUTO. My analysis focuses on the weather normalised values — the energy or quantity that building would have used under average conditions.


Figure 1

As mentioned earlier, not all buildings are subject to LL84. Square footage is what implicates a building, so naturally there are certain types of use properties that we would imagine to be most prevalent in this dataset — offices, schools, and hotels among them (Figure 1).

We can see that a disproportionate percent of total benchmarked floor area comes from buildings larger than 500,000 square feet, though the most benchmarked buildings were smaller than 100,000 square feet (Figure 2, Figure 3).

Figure 2, Figure 3

Source energy use intensity is calculated by the Portfolio Manager at the source of energy generation in kBtus per gross square foot (kBtu/ft2) for the reporting year. As discussed earlier, this value is normalised for weather. Figure 4 illustrates that buildings with more floor area, built in the earlier part of the last century, have higher median source energy use intensity. While potentially interesting, this visualisation is very likely to be skewed due to outliers. The generally high median EUI in large buildings doesn’t explain the incredibly high median EUI measured in a building built in the 1990s. Properties built in the 2010s have the lowest median EUI, consistent with the expectation that newer construction is built more energy efficiently.

Figure 4
Figure 5

Multifamily buildings far outweigh the rest of the sample (Figure 1), both in count of buildings and in gross floor area subject to benchmarking. Perhaps unsurprisingly, Median emissions per unit decline steadily when looking across years that properties were built (Figure 5).

Next Steps

This analysis didn’t address two important questions. For one, there’s an overarching question of what actually incentivises owners to improve their buildings’ energy performance. Just as Bloomberg claimed credit for a reduction in energy consumption that likely would’ve occurred whether or not he was in office, it’s not clear whether the measures undertaken by the deBlasio administration are simply making public improvements that would’ve otherwise taken place.

Further analysis may look more closely at changes in energy consumption of properties over time. One could conceivably track properties that have undergone extensive retrofits. When compared to buildings that have undergone less extensive energy efficiency improvements, one might evaluate the range of options available to property owners and identify those most optimal.

Second, Local Law 97 (LL97), the recent measure passed as part of the City’s Climate Mobilization Act, is set to achieve an overall reduction in greenhouse gas emissions of 80 percent by 2050 compared to 2005 levels. As was discussed in my analysis, multifamily housing properties are a big contributor to overall building energy consumption in the City. However under LL97, rent regulated units are exempt — not subject to the carbon emissions caps in the same way that the rest of the stock is.

Additional analysis can look closely at the proportion of rent regulated units reported on under LL84. Are there sufficient incentives or mechanisms for owners of the regulated stock to make energy efficient upgrades to those units? More ambitiously, is New York City capable of taking on a Green New Deal for Public Housing, in which public housing developments undergo extensive retrofits while creating jobs for New Yorkers?

The code and data used in this analysis can be found on GitHub.