A $50 million assessment, and the one question that comes first: who controls the thermostat.

The decision, the way the co-op is being told it

The one question that governs the decision

Every figure in the pitch answers the same buried question: the buildings are inefficient, so will electrifying them fix it? For an apartment building, the framework asks a different one, and this is important: it is not something we discovered about your building. It is the question the framework asks of every building of this kind, before it looks at any number at all. Think of it as the prior an experienced underwriter brings to the room:

Does the owner actually control the load it is about to be fined for?

In plain terms: in a building of 2,904 apartments occupied by individual families, who controls the thermostat? If a shareholder in unit 4C runs the heat at 78 degrees with the window cracked, that is fuel burned, carbon emitted, and, under Local Law 97, penalty owed. The co-op signs the check. But the co-op did not turn the dial. That distinction is the whole game, because a $50 million capital program only works if spending the money actually moves the number the city fines you on. Hold that question. Everything below is what the public record can, and cannot, say about it.

What the public record actually shows

New York has required each Glen Oaks section to report its measured energy use every year since 2010, under Local Law 84. That is not an engineer's model. It is the meter reading, filed with the city.

The bounded asset: Section O, fourteen years of its own meter

The framework needs one identifiable asset to reason about. We used Section O (71-33 260th Street), the section with the highest reported energy intensity, and the one carrying its own Local Law 97 exposure. Here is its full filed history.

¹ Site EUI (kBtu/ft²): the energy a building uses per square foot per year, its “miles per gallon.” Lower is better. A normal New York apartment building runs somewhere between 50 and 200. ² Weather-normalized: the city's own correction for hot and cold years, so you compare the building to itself, not to the weather. ³ ENERGY STAR (1–100): where this building ranks against similar buildings nationwide. A score of 4 means roughly 96% of comparable buildings do better; a score of 97 means only 3% do.

Weather-normalized means the city already removed the weather, so this swing is real, not “it was a cold winter.” Over fourteen years the series has no upward trend and no downward trend. It is unstable, moving between 144.8 and 213.9 and landing in 2024 essentially where it was a decade earlier. A $50 million program cannot be sized against a number that will not hold still.

A careful note, because it is easy to get this wrong: an earlier reading of a shorter slice of this series called Section O “deteriorating.” That was drawn from three years and is withdrawn. Over the full record it is not deteriorating and not improving. It is unstable.

A uniform solution against a non-uniform condition

Now widen the lens from one section to the whole development. In calendar year 2020, the Glen Oaks sections that filed plausible numbers reported this.

Read the top and bottom bars together. Section D reports 46.1 with an ENERGY STAR score of 97, outperforming 96% of comparable multifamily buildings in the country. Section O reports 152.2 with a score of 4. Same 1947 vintage, same self-management, same 110 acres, same weather: a 3.3x spread in a single year. This does not prove the buildings are fine, but if 1947 vintage were the driver, buildings of identical vintage would not land 3.3x apart, so 1947 cannot be the whole story. The crux the framework will not let pass: a uniform capital solution is proposed against a non-uniform measured condition.

The honest caveat: the measurement itself is shaky

We will not overclaim the dispersion, because the public record has a hygiene problem the board should know about. Weather-normalized values are published for only 3 of the 13 reporting sections, so the cleanest apples-to-apples comparison exists for a minority of the campus. Where those normalized values do exist, the spread survives: 69.5 to 160.6, a 2.31x range.

The series is also internally inconsistent to the point of impossibility. Nine section-year records carry site EUI values above 1,000 kBtu/ft², against a normal multifamily range of 50 to 200. Section B is reported at 14,316 kBtu/ft² in 2015, physically impossible, and it sat in the published record uncorrected for years. And the “best” section is not stable either: Section D, the ENERGY STAR 97 standout, collapses from the 97th percentile in 2020 to the 22nd by 2024, with no reported physical change. Any thesis built on a single year of this data is built on noise.

The strongest defensible statement is not “the buildings are efficient” and not “the buildings are inefficient.” It is this: the city's own mandatory disclosure for Glen Oaks cannot presently tell you which building has a problem, or how big it is. It is non-uniform across sections and unstable within them. That is precisely the wrong foundation on which to size a $50 million, one-size-fits-all capital program.

The arithmetic the board is actually voting on

Two formulas are worth putting on the table, because they are where a round number meets a real invoice. Local Law 97 fines a building on a simple line.

Notice what the formula needs that we do not have in the public record: your cap. The figures quoted in the press, $394,000 a year starting in 2024, rising to $1.5 million a year by 2030, are projections attributed to the cooperative, not penalties assessed in any public DOB record we could locate. We flag that as a gap, not a fact. For scale only: Section O's reported emissions in 2020 were 1,339.6 tons, but without the section's assigned cap the penalty cannot be computed from public data.

The assessment's own headline numbers do not quite reconcile either.

No public engineering study, and no Local Law 87 audit report tying this scope to measured per-section conditions, was located in this sweep. The scope exists in press reporting of board statements. It does not yet exist in a technical document anyone can check.

Why the honest answer is “not yet”

Most analysis stops here with a shrug, “hard to say without more data.” A governed read does the opposite. It states exactly how much it knows, on a fixed ladder, and exactly what would move it up.

The energy number reaches L3, it is the city's mandated measurement, filed for fourteen years. But a decision does not run on one number. It runs on a cluster of variables, and here the framework applies a rule any engineer, and any co-op treasurer, already knows.

A chain is only as strong as its weakest link. The framework calls it the meet: a cluster cannot be graded higher than its weakest member.

To decide whether to spend $50 million avoiding a carbon penalty, you need the measured energy (have it, L3), the cause of the dispersion between sections (not observed, L0), the heating and control condition per section (L0), whether each section's meter boundary is even physical or just an administrative split (L0), and the governing one, who controls the load being penalized (L0). One link is at L3. The rest are on the floor. So the cluster is on the floor, and the framework's evidence ceiling for this decision is NOT_OBSERVED.

The verdict is a Decision-Blocked Asset Brief. Not “no,” and not “the penalty is fake.” It is, “not on this evidence, and here is the evidence that changes it.” Concretely, the run put all 8 open decision cases into investigate, zero into act, zero into defer, and of the actions it weighed it permitted exactly one: request the missing evidence from the operator. The client's own thesis, 1947 buildings are old so electrify, the framework logged as a hypothesis to inspect (LC-OPS-02), supported for screening only, not for a decision.

New York already made you produce the evidence that lifts this

Here is the quietly useful part. The same stack of city laws that created your penalty also required you to produce most of the evidence that would let you size the response. It is sitting in a compliance folder. But read the residential fine print carefully, because it is itself part of the finding.

LL87: buildings over 50,000 ft² must complete an energy audit and retro-commissioning every 10 years. Section O (156,259 ft²) qualifies, and that audit is the “why” behind the EUI. LL88: requires sub-meters and monthly statements in tenant and non-residential space over 5,000 ft², commercial units and community rooms, not in individual apartments. LL97: penalty = (reported emissions − your cap) × $268 per ton, every year over the limit. Your cap is the number the public record is missing.

The residential twist is the honest part of the finding. In an office tower, LL88 sub-metering nearly hands you the control-boundary evidence the framework is asking for: every major tenant load is metered and billed. In a residential co-op, it does not. LL88 reaches your non-residential space, not the 2,904 apartments where the heat is actually consumed. Which means the co-op may genuinely not control, or even separately measure, the load it is about to be fined for. That is the single most important fact for a $50 million vote, because electrifying a load the resident still controls does not automatically move the emissions the city bills you on the way the model assumes.