Showcase · Data layers · NYC Open Data

Which buildings face Local Law 97 fines?

Which New York City buildings are over their Local Law 97 limit, and what would they pay? Each benchmarked building’s emissions against its limit, from the city’s own energy data for 2024.

Local Law 97 of 2019 caps the greenhouse gas emissions of most buildings over 25,000 square feet (2,323 m²). From 2024 a building over its cap pays $268 for every tonne of CO2e over, each year, and the caps tighten in 2030. On what they used in calendar 2024, 1,830 of 21,753 benchmarked properties (8.4%) are over their 2024 to 2029 limit: about $58.6 million a year. On the same energy use, 13,546 (62.3%) would be over the 2030 to 2034 limit, about $532.8 million a year.

The city publishes the energy, not the limit

Every year the owners of large buildings report their energy and water use under Local Law 84, and the city publishes it: 39,090 records for calendar 2024, each with its fuel use and its floor area by use. What the table does not have is the building’s Local Law 97 limit, its emissions as the law counts them, or anything to map it by. The limit comes from a rule of the Department of Buildings (1 RCNY 103-14): each use’s floor area times a factor in tonnes of CO2e per square foot, 60 property types in all, one table for 2024 to 2029 and a much tighter one for 2030 to 2034. The emissions come from each fuel times its coefficient in the law. Here both were worked out for every record, then each property was placed on its tax lot (PLUTO) and keyed to the map in Plotlay.

The map

The 59 community districts, coloured by the likely penalty or by the share of properties over the limit, or Midtown’s buildings one by one. Switch the period. Hover over a shape, or tap it on a phone, for its numbers.

2030 is the cliff

Most buildings are under their limit today. 11,719 properties that are under it in 2024 to 2029 are over it in 2030 to 2034 if they change nothing. The 2030 factors are roughly half the 2024 ones or less, and the law’s coefficient for grid electricity falls too (from 0.000288962 to 0.000145 tCO2e per kWh), which helps buildings that run on electricity and does nothing for gas and oil. Multifamily housing goes from 8.6% over to 62.7%, offices from 5.0% to 53.2%.

Manhattan carries most of the bill. The five districts with the largest 2030 to 2034 totals, Manhattan 5, Manhattan 12, Manhattan 8, Manhattan 2, Manhattan 7, add up to 33.7% of the city’s.

Midtown, building by building

A district total hides the buildings. In Manhattan Community District 5, which runs from 14th to 59th Street, roughly between Eighth and Lexington Avenues, and holds much of Midtown’s office space, Plotlay took every footprint of every benchmarked property from the city’s footprints file at once and keyed each to its Local Law 84 property id: 1,530 building outlines for 1,431 properties, the buildings of one property joined as parts of one key. In 2024 to 2029, 113 of the district’s 1,507 properties in the totals are over, about $4.8 million a year. In 2030 to 2034 it is 869, about $62.1 million, 11.7% of the city’s total from one district.

The total depends on what you count

Not every building on the list pays. City-owned buildings, NYCHA, HDFC co-ops, houses of worship and income-restricted housing follow other compliance routes, and some records do not add up. Here they are flagged and kept out of the totals, so the headline is a careful figure, not the largest one. Putting back the 1,204 records that fail a data check raises 2024 to 2029 to $118.8M; every pathway as well gives $265.0M.

How it was made in Plotlay

1. A map project on the USGS National Map topo base map (public domain, new in Plotlay for the United States), and two data layers.
2. Community districts: Take all. The City Planning districts file added as a layer to pick from, with its licence, then every area at once, keyed by its own boro_cd (101 to 503).
3. Place rows by their position. The 21,753 properties in the totals, each at its PLUTO latitude and longitude, fall into the district they stand in. Both periods are rows of one table, so the period is the time axis and the slider steps between them.
4. Colour by the share over the limit (the average of a 0 or 1 column) and by the likely penalty (the total), breaks set by hand.
5. Midtown buildings: Take all from the city’s footprints file, keyed by the LL84 property_id each footprint was matched to (by its BIN, else its lot).
6. Link by key the Midtown table, colour by the likely penalty on the period slider, and leave out the properties on another pathway or failing a check, in grey with their own legend row.
7. 3D at the real heights: each footprint at its roof height in metres (the city publishes feet).
8. Charts and map pictures in the Analysis workspace: the maps on this page, the share over the limit by kind of building, and what the total depends on, each exported as PNG.
9. Export GeoJSON, TopoJSON for the Power BI Shape Map, CSV with WKT and the linked data.

More about Data layers · Make a data layer of your own

Power BI

Python

A narrated Jupyter notebook reads the tables, works one building’s limit out by hand from its uses and the rule’s factors, accounts for every record, recomputes every number on this page, and maps the districts and Midtown with GeoPandas.

Download the notebook (with its outputs) and the data. Both are also in the plotlay-nyc-ll97 repository on GitHub.

Read it carefully

• An estimate, not a bill. It is what a building would pay at $268 a tonne if its 2024 report stood as published. Owners can correct a report, use the good faith provisions for 2024 and 2025, buy renewable energy credits or apply for an adjustment, and none of that shows in open data.
• 2030 to 2034 assumes nothing changes: the 2030 limits and coefficients applied to calendar 2024 energy use.
• Matching records to lots. 27,048 of 27,270 kept records matched a PLUTO lot by their BBL; 222 did not and have no district. 99.9% of the PLUTO positions fall inside PLUTO’s own district for the lot. In Midtown, 150 of 1,581 properties have no footprint of their own, most because they share a building with a larger record.
• Mixed use. 32.2% of the properties in the totals list more than one use; the limit adds each use’s floor area times its factor, as the rule says. 930 have a use under a newer Portfolio Manager name or one the rule does not list, which takes the rule’s successor name or its “Other” factor for the same category (our assumption, listed in the data).
• Pathways are flags, not facts. 4,473 properties are likely on another route and kept out of the totals: city-owned 1,737, tax-exempt housing 1,163, HDFC 942, NYCHA 237, state or federal 224, houses of worship 170, from PLUTO’s owner type and name. Rent-regulated buildings (more than 35% of units) follow Article 321 too and cannot be told apart in open data, so some multifamily properties in the totals will not face these penalties.
• Data checks. 1,204 records whose energy per square foot is far from their property type’s median, or whose emissions disagree with Portfolio Manager’s own figure, are kept in the tables and out of the totals. 62.9% of the properties in the totals carry an estimated or default value somewhere in their record.
• Left out. 9,988 child records whose parent record already holds their energy, 1,291 with no energy reported and 209 with implausible energy per square foot, among others; each is listed with its reason.

More energy and buildings showcases

• You can’t manage what you can’t map: where a council uses its electricity, gas and water, from its own bills.
• One school campus, building by building: footprints keyed to a register, floor space estimated from heights, in 3D.