NEXA V1 — References & Methodology

A — What is NEXA

NEXA stands for Neighbourhood Energy eXploration & Analysis. It is an interactive neighbourhood energy explorer. The user composes a neighbourhood by typology (Layer 1), then overlays energy demand + generation systems (Layer 2), electric mobility (Layer 3), and green infrastructure (Layer 4). Behind each neighbourhood archetype sits a full EnergyPlus simulation campaign across a cumulative ladder of energy-efficiency scenarios, so the user can compare a code baseline against each successive measure.

What NEXA V1 is, and what it is not yet

This release is NEXA V1. Everything described on this page is what V1 contains. Anything not listed here is not in V1, and later work becomes V2.

Intended use

NEXA is an early stage comparison tool. It is meant for a planner, a municipality or a researcher who wants to see how neighbourhood form, envelope standard, on-site generation and electric mobility move a small set of energy indicators, before any design work begins.

It is not, in this version:

• a design or compliance tool for an individual building,
• an optimisation engine: no formal optimisation method is implemented, so the tool never ranks the scenarios and never recommends one of them,
• a substitute for a project specific energy model.

Coverage, and what each status word means

Five climate arms are published in NEXA V1: Vancouver (zone 4), Toronto (zone 5), Montréal (zone 6), Winnipeg (zone 7A), Fort McMurray (zone 7B), each on NECB 2017. Each carries 35 neighbourhood units and a five rung scenario ladder, and each has a high performance twin. Montréal additionally carries the 1983 Quebec reference envelope. The exceptions to that coverage are listed in the limitations chapter and are not silent.

Two further arms were simulated and are not published in V1. NECB Zone 8 (Chisasibi) remains part of the research database and of the technical register in section J, and its quantitative results are not exposed in the public tool at this stage. The ASHRAE 90.1 arm is withdrawn from the climate selection: it is a United States reference case, run with United States prototypes on a United States weather file, and it is not a Canadian climate.

Four words, and no others, describe the maturity of a number anywhere in this tool:

• Simulation-backed: the figure comes from the EnergyPlus campaign. Layer 1 and Layer 2 results are in this class.
• Preliminary: the figure comes from a documented calculation chain rather than from a simulation. Layer 3 mobility and Layer 4 landscape PV are in this class, and both pages say so.
• In development: being worked on, no figure shown.
• Not modelled yet: no quantitative method exists, the option cannot be selected, and no figure is shown for it anywhere, including the exported scenario.

B — How the Four Layers Work

• Layer 1 — Neighbourhood selection: choose land use, context, density, layout, and envelope code. Returns matching neighbourhood archetypes.
• Layer 2 — Energy: energy demand systems (Heat Pump COP 4, Heat Pump for DHW, appliances) and on-site generation (PV). Shows EUI breakdown + PV generation.
• Layer 3 — Mobility (preliminary): EV demand and V2G balance.
• Layer 4 — Green infrastructure (preliminary): landscape PV (LPV) and green systems.

C — PV Methodology & the kWh/m² Definition

Two PV tracks

• Baseline: native PV only — whatever the source prototype ships with. ≈ 0 kWh/m² for Canadian baselines; a few high-rise/ASHRAE prototypes carry a small native array (e.g. 1.9–3.3 kWh/m²).
• HPerf and every rung above it: a “Tier-3” injector adds rooftop PV sized from the actual roof geometry.

Sizing rules

• Flat roofs → a single world-south rack at 45° tilt; active area = Σ(roof area) × GCR / cos(45°). The rack is corrected for the building’s North Axis so it always faces true south.
• Pitched roofs (houses, some small commercial) → panels on the south-facing dominant pitched face only (dominant-face fallback if no south slope); array tilt = roof pitch.

Assumptions: one panel, two mountings, and two figures that are not ours

Preliminary. Every photovoltaic number published in this tool was produced by the Tier 3 injector at a 23.0% panel efficiency. The array assumptions are a property of the roof, not of the neighbourhood, and what changes between the two roof groups is the mounting and the fraction of the surface that carries cells, not the panel. Common to both: DC→AC system losses 14%, inverter efficiency 96%, azimuth = world south.

Defect DBG-037, closed 2026-08-24. What changed, and what did not. Until this date the tool published 18.65% and called it the module efficiency the pipeline uses. That sentence was not true: 18.65% belongs to the Tier 1 track, which was retired on 2026-05-29, and the published results were produced by Tier 3 at 23.0%. No simulation was re-run and no energy figure moved. The generated kWh are unchanged EnergyPlus output; only the assumption printed beside them was corrected. Harmonising the three conventions into one upstream figure would change published results and is left to a later version, which is why this whole block is marked Preliminary.

The total PV array area, and why nine neighbourhoods do not show one. The array area is not an EnergyPlus result and it is in no results file. It is read back out of the simulation input: every photovoltaic surface injected on the EEM4 run of a neighbourhood, summed, each roof face counted at the active fraction the injector gave it, 0.85 on a pitched roof and 1.0 on the flat rack. A number is published only where that array reproduces the generation already on the page, the stored PV intensity times the conditioned area, to within 2%. Twenty-six of the thirty-five neighbourhoods pass and carry an area. Nine do not and the row is simply absent for them, rather than filled with an estimate: an array that does not reproduce the generation is not the array the printed generation came from. The nine are all pitched-roof house, townhouse and low-rise forms and all fall short in the same direction, while every flat-roof form reproduces to within 0.35%, which points at the south-facing-slope restriction the pitched-roof rule acquired when Tier 1 was retired. It is logged as DBG-038 and it is open. The area does not change with the climate: the injected array is identical to the decimal in Montréal, Vancouver and Fort McMurray, so it is stored once per neighbourhood.

Decisions D0.2 and D4.3, and defect DBG-013, closed 2026-08-10. A fourth figure, 18.68%, was stored on all thirty-five neighbourhoods and appears in no upstream source; it was deleted. Separately, the mounting and ground-coverage rows had been reporting the flat-roof set for all thirty-five, so the four house neighbourhoods were described as sitting on racks they do not have. Both roof sets now live once in js/config.js and every page reads them from there.

Units & normalization

PV intensity is reported as kWh per m² of heated and cooled floor area, the same denominator as EUI, so that the two can be compared and divided by one another. It is not normalized by roof area, and not by footprint. Section G defines that area exactly and tabulates it against the total building area.

Why it varies from one neighbourhood to another: the numerator (roof PV generation) is limited by available roof area, while the denominator grows with every storey. Low-rise forms (detached, row houses) have a large roof-area-per-floor-area ratio → high intensity; tall forms (mid-/high-rise) spread the same rooftop over many storeys → low intensity. This is physically expected, not an inconsistency. Representative NECB-2017 values: Detached (RC-D) ≈ 100.7, MidRise ≈ 66 to 75, HighRise (RC-HR1) ≈ 45.8, (RC-HR2) ≈ 26.0 kWh/m².

Total vs intensity are linked by: Total PV (kWh/yr) = PV intensity (kWh/m²) × heated and cooled floor area (m²). The worked example on record for RC-D is 100.7 × 10,560 = 1,063.4 MWh/yr, and 10,560 m² is the heated and cooled area, not the 21,201 m² total building area.

Corrected 2026-08-10, tracked as DBG-029. The absolute total on the PV page had been computed against the total building area rather than the heated and cooled area, so for the house-based neighbourhoods it read about twice the figure quoted above. The intensity and the ratio of PV to demand were never affected, because the same area cancels from both sides. The heated and cooled area is now carried for all thirty-five neighbourhoods, taken from the same simulation campaign that produced the energy results, so the identity above now holds on every page.

D — Ratio of Performance (RoP)

RoP is not a website invention. It is a published metric of the neighbourhood paper, which defines it as the ratio of annual on-site PV generation to annual total building energy demand, site energy, per Neighbourhood Unit.

In plain words: what share of the energy this neighbourhood uses in a year can its own rooftop solar produce in that same year?

RoP = annual PV generation (kWh/m²) ÷ annual energy demand (kWh/m²)

Both quantities are annual, both are per square metre of heated and cooled floor area, and both are read from the same simulated scenario. The floor area is therefore on both sides of the division and cancels out, which is why RoP does not depend on which floor area is used. It is a pure ratio with no unit.

How to read the number

• RoP = 1.00 — over a year, rooftop PV generates as much energy as the neighbourhood consumes. This is annual net zero on a site-energy basis.
• RoP below 1.00 — the roof covers only part of the demand. An RoP of 0.35 means 35 % of the annual demand.
• RoP above 1.00 — the roof generates a yearly surplus.

What RoP does not say

RoP is an annual balance, not an hourly one. A neighbourhood at RoP 1.00 is still importing electricity on a December evening and exporting it at midday in June. Reaching net zero over the year does not mean the neighbourhood is independent of the grid, and RoP says nothing about storage, peak demand or the capacity of the local network. Layer 3 vehicle-to-grid and any battery are outside this ratio.

It is also a ratio of two intensities, so a low RoP is often a statement about form rather than about the solar design. A high-rise has roughly the same roof as a low-rise but many times the floor area beneath it, so its RoP is structurally lower even with an identical PV system. Section C explains that effect.

Worked example

RC-D, detached houses, NECB 2017. At HPerf the PV generation is 100.4 kWh/m²·yr against a demand of 107.0 kWh/m²·yr, so RoP = 100.4 ÷ 107.0 = 0.94: the roof covers about 94 % of the year. Climb to the top rung, HPerf + Space HP + HPWH + EEM, and the roof is unchanged while demand falls to 60.6, giving RoP = 1.66. This is the general shape of the result: RoP rises because the measures cut demand, not because more panels were added.

On the baseline the same neighbourhood shows RoP = 0. That is correct, not missing data. The baseline models only the PV the building already had, and a detached house prototype ships with none. Section C explains that distinction.

RoP is computed when the page is drawn, from the PV intensity and the demand intensity of the scenario on screen. It is not a stored number, so it cannot fall out of step with the two values it is derived from. Before 2026-08-10 the tool carried a stored rop field that was empty for all thirty-five neighbourhoods; that field is gone.

E — Standards & Codes Implemented

Canadian baselines also apply BTAP (Building Technology Assessment Platform, NRCan/CanmetENERGY) internal-load densities and schedules: occupancy 0.04 people/m², plug loads 7.5 W/m², NECB thermostat schedules.

NECB 2017 prescriptive caps used (selected): wall U ≤ 0.247 (Z6) / 0.210 (Z7A) W/m²·K; roof U ≤ 0.183 / 0.162; window U ≤ 1.90 / 1.60; air leakage 0.25 L/(s·m²) @ 75 Pa; HVAC efficiency floors (DX cooling COP ≥ 3.0, DX heating COP ≥ 3.1, gas/boiler ≥ 0.80, chiller COP ≥ 4.7). NBC 9.36 residential: window U ≤ 1.60 (Z6) / 1.40 (Z7A); air leakage 2.5 ACH₅₀ (Z6) / 2.0 ACH₅₀ (Z7A).

F — Climate Zone & Weather

Seven climate arms were simulated. Five are published in NEXA V1. The ASHRAE arm is a United States reference case and Zone 8, Chisasibi, is held back at the supervisor's request; both remain in the research database. Each climate the tool offers is a separate EnergyPlus campaign on its own weather file. The exact file is named on every result page, under Assumptions & Model Information.

G — Building Archetypes & Sources

• Commercial: 16 DOE/PNNL ASHRAE 90.1-2022 prototypes (offices, retail, schools, hotels, hospital, outpatient, restaurants, warehouse, apartment mid/high-rise) + 2 in-house high-rise variants (15, 20 storeys).
• Residential: 3 IECC 2024 single-family variants (detached + garage, no-garage, row-house).
• Supplemental (OpenStudio, 90.1-2019): College, Laboratory, TallBuilding, SuperTallBuilding, 2 small datacenters.
• Datacenters: ASHRAE 90.1-2019 large-format prototypes (passed through, NECB-exempt).
• Three large prototypes (Primary School, Secondary School, Warehouse) are used at 50% floor area to fit neighbourhood plots (see Resizing).

H — Energy Modelling Basis

• Engine: EnergyPlus 22.1 (mid/high-rise + commercial) and 23.1 (single-family/attached).
• EUI: site energy use intensity in kWh/m²·yr, per heated and cooled floor area. PV generation is reported separately and never netted into EUI. Both terms are defined exactly below.
• Neighbourhood airflow: AirflowNetwork:SimulationControl = MultizoneWithoutDistribution.
• Envelope-code options offered per neighbourhood: Standard, the NECB 2017 baseline for the selected climate zone, and HPerf, the High-Performance Envelope. HPerf is an envelope package only: it is applied strictly to the building envelope of a building built to NECB 2017, and it is the first rung of the scenario ladder. Montréal additionally offers a 1983 Quebec reference envelope starting point. The ASHRAE 90.1 options were withdrawn from the selection on 17 August 2026; see section F.

The reference case, and why nothing is compared until you choose one

Until 24 August 2026 the results pages drew a column headed Baseline without being asked. That column is the same neighbourhood, in the same climate, built to the NECB 2017 code minimum with no measures applied. The arithmetic was right and the name was not: a column called Baseline reads as the neighbourhood as it exists today, and it is nothing of the kind. Nothing in this tool has ever looked at how the buildings in a neighbourhood are actually built today.

In NEXA V1 the reference case is chosen by the reader, and nothing is compared until it is. Two references exist:

• NECB 2017 code minimum. The same neighbourhood, in the same climate, at the code minimum with no measures applied. Available in every published climate.
• 1983 Quebec reference envelope. The same neighbourhood built to Quebec Regulation E-1.1, r. 1, Order in Council 89-83. It is the earlier code era, and it was simulated for Montréal only, so it is offered there and nowhere else. See section M.

Both are code-era models, not surveys of an existing building stock. A comparison against either one is a comparison between two modelled code editions of the same neighbourhood, and that is what the pages say when a reference is chosen.

How the EUI is computed

EUI (kWh/m²·yr) = Total End Uses (kWh/yr) ÷ Net Conditioned Building Area (m²)

• Numerator. The EnergyPlus Total End Uses row of the Annual Building Utility Performance Summary: electricity plus natural gas, with no source-energy multipliers applied. It is site energy, not source energy. It covers every end use the model reports, including the ones the breakdown chart does not draw separately. Those are collected into the block labelled Other, so the chart adds up to the number printed above it.
• Denominator. The EnergyPlus Net Conditioned Building Area: the floor area the model actually heats and cools. This is what the pages mean by heated and cooled floor area.
• Not included. On-site PV generation is never subtracted from the EUI. It is reported separately, in section C.

How natural gas is converted to kWh before it is added to electricity

Both fuels are read from the same EnergyPlus End Uses table, in the same reported unit, which is gigajoules in these runs, and both go through the same conversion:

1 GJ = 277.778 kWh, which is the exact identity 1 kWh = 3.6 MJ. Where the table is reported in kBtu instead, the factor is 1 kBtu = 0.293071 kWh.

Nothing is weighted by fuel. A kWh of gas and a kWh of electricity are added as equals, because this is site energy. No source-energy or primary-energy conversion factor, no grid emissions factor and no fuel-specific multiplier is applied at any point in the chain. A tool that reported source energy would multiply electricity by roughly 1.8 to 3 and gas by roughly 1.05, depending on the jurisdiction; this one does neither.

The two floor areas, and which number uses which

EnergyPlus reports two building areas, and this tool uses both, for two different jobs. Total Building Area is everything enclosed, including unheated attics and unconditioned basements; it is the floor area listed on the neighbourhood table. Net Conditioned Building Area is the heated and cooled part of it, and it is the denominator of every intensity on the energy pages.

The gap is not small. For the neighbourhoods built from houses the conditioned share is exactly one half, because the attic and the unheated basement together are half the enclosed area and neither is conditioned. For the apartment neighbourhoods the share is close to nine tenths.

Decision D6.0, 2026-08-10: the published intensities stay on the heated and cooled area, which is the basis the simulation wrote, and every page that carries an intensity now names that basis in words. No stored value was changed. The nine areas above are measured, not assumed, and are read from the EnergyPlus Building Area table for each neighbourhood; the total-area column matches the floor areas this tool ships. The remaining twenty-six neighbourhoods have not had their conditioned area tabulated yet, so this table lists the nine that have.

Scenario ladder (cumulative)

The five rungs are cumulative: each one keeps everything below it and adds one measure package. These names are the only ones used anywhere in the tool. If you open the raw data file or the source CSVs, the six rows above are stored in that same order under the internal keys DEFAULT, EEM1 to EEM4 and IAL; those keys are file column values, not names for the scenarios, and they appear nowhere on any page.

I — Layer 3: Mobility (EV / V2G) PRELIMINARY

Assumptions, from Templates/NUS_EV.csv, applied to all 35 neighbourhood units. Names and units follow the RT01 terminology review and are held once, in js/config.js, so the page and this chapter cannot drift apart.

• EV ownership rate: 1.5 EVs per household. This is an ownership rate, not a penetration rate: a penetration rate is the share of a vehicle stock that is electric, expressed as a percentage.
• Specific daily charging demand: 15 kWh per EV per day, about 200 km of range.
• Grid to vehicle charging efficiency: 90 %.
• V2G fleet participation rate: 50 % of the EVs.
• Specific daily V2G discharge energy: 10 kWh per participating EV per day. This was published as "10 kW / day" until 2026-08-12. That is not a valid unit: kW is a power and kW per day is neither a power nor an energy. The quantity was always a daily energy.
• Vehicle to grid discharging efficiency: 90 %. It is displayed for reference and is not applied to the exported energy in this version. The method is under review, so the figure below assumes the full 10 kWh reaches the grid.

The calculation chain

Per neighbourhood, per day. Worked example: RC-R, 24 households.

1. Total EV charging demand = households x 1.5 EV per household x 15 kWh per EV / 0.90 charging efficiency.
   24 x 1.5 x 15 / 0.90 = 600 kWh/day.
2. Storage loss = 5 % of the charging demand.
   600 x 0.05 = 30 kWh/day. The origin of this 5 % is undocumented upstream and is an open question with the supervisor, listed in the limitations chapter.
3. Daily V2G energy delivered = households x 1.5 EV x 0.50 participation x 10 kWh per participating EV.
   24 x 1.5 x 0.50 x 10 = 180 kWh/day. Zero when V2G is not selected.
4. Net grid demand (total) = charging demand + storage loss, minus the V2G energy delivered.
   600 + 30 = 630 kWh/day without V2G; 630 - 180 = 450 kWh/day with it.
5. Net grid demand intensity = net grid demand / heated and cooled floor area.
   450 / 5,300 m² = 0.08 kWh per m² per day. Until 2026-08-12 this was divided by a superseded floor area table and read up to three times too high on the tall neighbourhoods.

Sign convention

A positive net grid demand means the neighbourhood must draw that much energy from the grid each day, which is a deficit. The tool follows the delivered energy convention of ISO 52000-1 and ASHRAE Standard 228. Note that the net zero building and positive energy district literature uses the opposite sign, where positive means a surplus, so the direction is stated in words next to every figure rather than left to the reader.

The status indicator has three states: grid import required (deficit), net zero balance, and net energy export (surplus). Only the first can occur today, because the on-site PV generation is not netted against the EV load. Whether it should be is a methodology question with the supervisor.

Status: Preliminary. These figures come from the fixed chain above, not from a building energy simulation.

J — Layer 4: Green Infrastructure / Landscape PV (LPV) PRELIMINARY

Landscape PV is the one Layer 4 option with a calculation method behind it, so it is the one option that produces a number. The chain is held once, in js/config.js, and every row on the Layer 4 page is generated from it.

The calculation chain

Identical for all 35 neighbourhood units, by decision. See the note below.

1. Land / site area: 5 acres, 20,234 m².
2. Land allocation: 20 % of the site = 4,046.8 m².
3. Usable area: 10 % of the allocated land = 404.7 m². The two fractions are applied in series. Until 2026-08-12 they were applied in parallel, so the usable area was reported as 2,023 m², which is 10 % of the whole site, and the 20 % allocation was silently discarded.
4. Installed capacity: 404.7 m² x 200 W/m², which is a 400 W module on 2 m², = 80.9 kWp. The figure published before 2026-08-12 was 475 kWp, which would require 29.3 acres of land rather than 5.
5. Annual generation: 80.9 kWp x 1,280 kWh/kWp per year = 103.6 MWh/yr. The specific yield of 1,280 kWh/kWp per year is sound for southern Québec and was never the problem; the capacity it multiplied was. The figure published before 2026-08-12 was 608 MWh/yr, which is 5.87 times the result its own configuration rows produce.

Why every neighbourhood shows the same number

A uniform 5 acre landscape allocation is applied to every neighbourhood unit and is not scaled to neighbourhood size. This is a project decision recorded upstream, not a defect: per-neighbourhood variation in landscape PV is intentionally absent in this version.

The 10 % usable fraction is below the published range for fixed tilt ground mount installations, which is roughly 30 to 50 %. Raising it would be a redesign of the landscape PV facility rather than a correction, and it belongs to a later round.

The other green infrastructure options

Green roofs, vertical greening systems, linear greenery, green spaces, roof gardens, food gardens and PV-green roof integrated modules have no quantitative method in this version. They are grouped on the page under "Future capabilities", cannot be selected, and produce no energy figure anywhere in the tool, including the final summary and the exported scenario.

Status: Preliminary. Landscape PV is a fixed area assumption multiplied by a specific yield, not a building energy simulation.

K — Resizing Methodology

Three oversized prototypes (Primary School, Secondary School, Warehouse) run at 50% floor area to fit the largest neighbourhood plots; geometry proportions and HVAC sizing ratios are preserved. High-rise storey variants (15/20) are derived by ZoneGroup multiplier with corrected zone Z-origins so solar and wind exposure stay physically correct.

L — Validation Summary

• Whole-building EUI ladder validated against deep-energy-retrofit literature (NRCan PEER/BTAP, PNNL, NEEA/WSU): HPerf + Space HP + HPWH ≈ −33% to −42%, HPerf + Space HP + HPWH + EEM ≈ −52% to −62% vs baseline — within literature bands for 19 of 22 measures (2 conservative/under-credit, 1 no-direct-evidence).
• Ideal-load (IAL) thermal demand validated against NRCan CEUD/NEUD and PNNL prototype baselines (heating, DHW, IAL/DEFAULT ratios all within expected ranges).

M — 1983 Quebec Reference Envelope (NECB Zone 6, Montréal)

What the 1983 model is, exactly. It is a code-era reference envelope, built to Quebec Regulation E-1.1, r. 1, Order in Council 89-83, the 1983 Quebec construction requirements. It is not a survey of how the buildings in a neighbourhood are actually built today: no postal-code or feeder-resolved consumption data exists for Montréal, so the arm was validated against national and provincial survey composites rather than against measured meters. The arm is CAN_MTL_1983, it covers all 35 neighbourhood units, and it exists for Montréal only. Source: the LMN-1983 chapter of the upstream documentation and its validation report, closed 2026-08-09.

The 1983 Quebec reference envelope represents the pre-code building stock typical of Montréal and Quebec’s existing urban fabric. These envelope parameters are based on construction practices prevalent before the adoption of modern energy codes, featuring significantly lower insulation levels, higher infiltration rates, and single- or double-pane glazing systems.

Purpose & Context

While the Standard (NECB 2017) and HPerf (High-Performance Envelope) options model new construction, the 1983 reference envelope captures the retrofit potential of existing buildings. By comparing 1983 baseline EUI against progressively deeper scenarios on the ladder above, users can estimate the energy savings achievable through envelope upgrades, HVAC electrification, DHW heat pumps, and efficient lighting/equipment on real existing building stock.

Simulation Campaign

• Weather file: Montréal (CWEC — CAN_MTL), Climate Zone 6A.
• Neighbourhood units: All 35 NUs (RC, RS, MU, CC, IC archetypes).
• Scenarios per NU: 16 upgrade-measure combinations (4 binary flags: Envelope, HVAC, DHW, Lighting/Equipment) — 560 total simulation runs. The five published rungs are the cumulative subset of those 16.
• Engine: EnergyPlus v22.1, consistent with the Standard and HPerf campaigns.

Key Envelope Differences vs. NECB 2017

EUI Comparison (Baseline)

The 1983 reference envelope at the Baseline rung (no upgrade measures) shows substantially higher energy use intensities than the NECB 2017 baseline, primarily driven by increased heating demand due to poor envelope thermal performance. Typical increases range from +40% to +80% in heating EUI depending on the archetype, with mixed-use and commercial typologies showing the largest absolute differences.

Validation & Documentation

A comprehensive validation report covering all 35 neighbourhood units is available:

Availability

The 1983 Quebec reference envelope is currently available only for NECB Zone 6 (Montréal). When selecting an envelope in Layer 1, the “1983 — Quebec reference” button appears exclusively in the NECB Zone 6 performance-level popup. All 35 neighbourhood archetypes are supported under this reference envelope.

N — Limitations

Envelope & HVAC Modeling Constraints

• AirflowNetwork Numerical Singularities: To prevent numerical singularities in the AirflowNetwork solver during high-wind winter events at the neighbourhood scale, the detailed duct distribution and leakage solver (`MultizoneWithDistribution`) is disabled. Neighbourhood simulations use the simpler `MultizoneWithoutDistribution` mode, dropping the duct leakage component (estimated to introduce a ≤ 5% EUI difference, but reducing simulation times tenfold).
• Daylighting and Venetian Blinds Skip Rules: Daylighting controls are automatically skipped in zones with continuous-use schedules (e.g., life-safety/process) or specialty functions (labs, kitchens, surgical rooms) where dimming is impractical. Automated interior venetian blinds are skipped on Laboratory, Tall, and SuperTall buildings due to EnergyPlus v22.1 engine crash constraints.
• Stratified Tank Reporting Anomaly (Phase P): In EnergyPlus v22.1, central stratified HPWH backup-resistance energy (≈ 7 kWh/m²·yr) is mis-bucketed under Space Heating instead of Water Heating. While the total combined energy is conserved, the individual end-use columns contain this reporting discrepancy.
• Process-Load Dominated Typologies: Industrial and IT-heavy archetypes (e.g., Data Centers, Laboratories) exhibit low EUI sensitivity to typical envelope and HVAC retrofits because their energy budgets are heavily dominated by IT plug loads and ventilation process exhaust, which are outside the scope of the building-level measures on the ladder.

Solar Photovoltaic (PV) Modeling Constraints

• Facade PV Validation: Facade PV (building-integrated photovoltaics on vertical walls) is only applied to high-rise typologies (ApartmentHighRise, ST15, and ST20) where vertical area is significant. Additionally, facade PV yields have only been simulated and validated for Montreal (CZ6A), with no second climate validated. The upstream facade study covered Montreal and Calgary, and Calgary is not one of the climates this tool offers.
• Pitched Commercial PV Sizing: Sizing values for pitched commercial buildings (OfficeSmall, RestaurantFastFood, RestaurantSitDown) are pending re-simulation under the south-facing dominant-pitched-roof constraint; current reporting reflects pre-fix values across all slopes.
• Racking Tilt vs. Market Standards: While flat-roof PV racks are modeled at an energy-optimal tilt of 45° (or 20° on the top rung, HPerf + Space HP + HPWH + EEM) for high-latitude winter performance, this tilt angle exceeds typical flat-roof commercial solar racking standards of 10–15°, which are constrained in practice by wind-uplift and structural ballast load limits.
• Snow-Cover Losses: Current PV generation calculations do not account for snow accumulation, snow-shedding, or winter albedo gains, which may overestimate actual winter yields in cold Canadian climates.
• No Electrical Storage or Grid Dynamics: The PV simulation layer assumes a simple baseload configuration where all generated DC electricity is inverted and fed directly into the building's electrical bus. No battery storage, peak-shaving control, or grid feed-in dynamics are modeled.

Coverage gaps, stated rather than hidden

These are cases where the tool either declines to show a result, or shows one that is known to rest on a weaker footing than the rest. They are listed here so that a reader never has to discover them from a number.

• Five neighbourhoods outside Montreal: RC-D, RC-ML, RC-MR1, RC-R and RC-T are merged from a United States detached house prototype in climate zones 4, 5, 7A, 7B and 8. The prototype is the zone 6A one in every case, so in those five climates these neighbourhoods differ from Montreal only by weather file. Montreal has a corrected campaign; the other five need re-simulation.
• Zone 7B, MU-HC: the baseline run for this pair does not exist upstream. The tool does not offer the combination and says so on the selection page, rather than filling the hole with a neighbouring value.
• The 5 % storage loss in Layer 3: its origin is undocumented. It may duplicate part of the 90 % charging efficiency, which already includes conversion losses. If it is instead the drain of a parked vehicle, the standards express that as a fixed amount per vehicle per day rather than a percentage of charging demand. Open with the supervisor.
• The 90 % discharging efficiency in Layer 3: displayed but not applied to the exported energy, so the V2G contribution is reported as if discharge were lossless. Open with the supervisor.
• On-site PV is not netted against the EV load: the mobility status indicator considers the EV load alone, so it can only ever report a deficit. Whether the two should be netted is a methodology decision, not a defect.
• Landscape PV usable fraction: 10 % of the allocated land, which is below the 30 to 50 % published range for fixed tilt ground mount. The result is conservative rather than wrong, and raising it is a redesign of the facility.
• Facade PV: offered only for the nine Montreal neighbourhoods that contain one of the three tall building archetypes it was tested on, marked Preliminary, and excluded from the totals.