Community comparison · 5 November 2026
How the new flare cases compare
The new inputs change the picture most clearly for prolonged HP upsets and gaseous emissions. They do not make every historical scenario worse: the old tank-shutdown case still produces substantially more particulate matter in these comparisons.
We compared the seven proposed 2026 source cases and three historical cases under five matched weather settings. The results below are source contributions at ten community reference points, with no background pollution added. They are conditional model results, not forecasts or measured exposure.
Download all community results (CSV)Download reproducible calculation packageMethods and data dictionary
Old and new cases under matched weather
| Weather and measure | Old compressor maintenance 30 minutes | Old tank shutdown 14 hours | New HP upset 24 hours |
|---|---|---|---|
| Inflow with inversion · 24-hour PM₂.₅ | 0.29 | 1.72 | 0.72 |
| Weak inflow with inversion · 24-hour PM₂.₅ | 2.81 | 16.43 | 4.04 |
| Weak inflow with inversion · highest hourly total VOC | 114.42 | 2.71 | 598.42 |
“Inflow” uses wind from 200° at 3 m/s; “weak inflow” uses the same direction at 0.5 m/s. Both have a prescribed inversion at 500 m above sea level. These are selected sensitivities, not a comprehensive worst-case search.
For the four pollutants reported in both inventories, the new HP upset has lower hourly rates than the old compressor-maintenance case, but holding its proposed maximum rate for 24 hours can produce larger daily contributions. Rates, duration and source geometry change between the defaults. The documents do not show that maximum hourly emissions persist for the full maximum event duration. 2026 notice, Table 2; 2014 inventory, Tables 22–23.
Across the study area
| Community reference point | Inflow + inversion 3 m/s | Weak inflow + inversion 0.5 m/s |
|---|---|---|
| Squamish | 0.11 | 1.52 |
| Brackendale | 0.72 | 3.62 |
| Britannia Beach | <0.001 | <0.001 |
| Garibaldi Highlands | 0.19 | 0.99 |
| Valleycliffe | 0.04 | 0.45 |
| Dentville | 0.23 | 2.24 |
| Garibaldi Estates | 0.30 | 1.99 |
| The Northyards | 0.64 | 4.04 |
| Eagle Run | 0.61 | 3.33 |
| Northridge | 0.08 | 0.82 |
“<0.001” is a very small calculated contribution, not a measured zero. Britannia Beach receives little contribution under these particular directions. Other directions and wind reversals were not bounded by this comparison.
None of the current cases crossed the Explorer’s numeric PM₂.₅ or SO₂ screening levels at the ten reference points in the five weather settings. Some near-source calculation cells did. Those flags did not intersect the mapped residential polygons for the current cases, but that mapping is incomplete. Point values and unshaded map areas do not establish safety or compliance.
What the gases tell us
In the weak-inflow HP-upset case, the largest calculated hourly total VOC contribution at a reference point is about 598 µg/m³. It is an unspeciated mass concentration: it cannot identify benzene exposure, a meaningful mixture-wide ppm value or illness risk. NOx is retained as NOx, rather than assumed to be ambient NO₂. The new vent case also includes H₂S and total reduced sulphur; those quantities can overlap and must not be added.
The proposed continuous sources are much smaller than the upset case, but their 24-hour calculations start with no earlier contribution from that source. This first-day window does not represent a long-running equilibrium or an annual exposure estimate. Source inputs, durations and citations.
How these comparisons were calculated
The exact Explorer worker and unchanged passive transport engine were replayed in 27 unit-source runs. Seven proposed source profiles and three historical profiles were then scaled by their documented or reconstructed g/s rates, producing 3,500 community/species rows.
- Five fixed wind settings: inversion inflow/outflow at 3 m/s, mixed inflow/outflow at 3 m/s, and weak inflow at 0.5 m/s. Inflow is from 200°, outflow from 20°. Upper wind follows the same direction. The inversion, when selected, is at 500 m above sea level.
- All calculations cover the first 24 hours. Venting represents only the first day of its 48-hour default release. Continuous sources use a clean-start 24-hour analysis window. No simultaneous HP + LP or whole-facility total is inferred.
- PM₂.₅ uses the full-day mean. Other constituents use the maximum of 24 one-hour means, which can occur at different times at different locations. PM removal is off.
- Historical injection is 200 m above the proxy’s model terrain, or 205.1552 m above sea level. The current physical-tip injection is 142.5 m above sea level at the application coordinates. This changes the map independently of the emission rates.
- The approximately 313 m calculation grid samples the lowest air layer above terrain. Layers are about 9–47 m deep at these reference points; values are not 2 m breathing-height predictions.
No flare buoyancy, exit momentum, atmospheric chemistry or background pollution is calculated. Numerical spreading remains a documented limitation. No weather probabilities, population weighting, illness calculation or reliable neighbourhood risk ranking follows. Transport benchmark.
The raw CSV retains unavailable rates as blank cells, not zeros. Its benchmark columns are numeric screening comparisons: PM₂.₅ 25 µg/m³ for 24 hours; SO₂ 170 µg/m³ for one hour; CO 14,300 µg/m³ as a one-hour legacy reference. They do not implement multi-year regulatory attainment. NOx and total VOC have no default benchmark in this comparison.
How the applicant’s assessment differs
The applicant’s separate CALMET/CALPUFF assessment reports results below provincial objectives at its assessed sensitive and Squamish town receptors, while identifying localized near-boundary NO₂ exceedances and potential odour exceedances during venting. These are different receptors, source combinations, physical treatments, weather periods and statistical measures from this Explorer comparison. They do not validate or directly contradict its individual numbers. TDR Tables 6.7–6.9, PDF pages 146–152.
The application’s venting maxima are near the facility boundary, not Squamish-town predictions. Its frequency estimates depend on assumed operating events and meteorology. The updated assessment did not carry VOCs/PAHs forward after the applicant judged their quantity/composition materially unchanged from the earlier EA; this comparison supplies no new individual-benzene exposure estimate. TAR discussion, PDF pages 82–84.
Check and reproduce the results
The download freezes the input HTML and source catalogue, exact engine/worker, recorded unit runs, full result tables and a portable rerunner. File-integrity and arithmetic checks are available without rerunning transport; a full offline replay uses Node.js and standard-library Python. These are AI-assisted software checks, not independent human scientific review or field validation.
- All 3,500 community/species rows (CSV) · Settings, provenance and full results (JSON)
- Reproducibility package · File hashes and package manifest · Recorded verification
- Methods, commands and data dictionary
This update adds a comparison of the existing Explorer cases. It does not change the simulator, adopt a new wind field or establish a prediction of local health effects.