Read the explanation
This page runs a synthetic eight hour queue model, not a live air traffic feed. Its stored JFK baseline is sixty operations per hour. A sixty five percent capacity reduction leaves thirty five percent of that baseline: twenty one per hour, or three point five per ten minute bucket. The engine permits fractional counts internally and rounds many display totals. After the outage, the authored capacity rule recovers linearly over one point five hours. During the outage, the ground delay input becomes a held fraction: buffer divided by one hundred twenty, capped at point eight five. In the first default bucket, eleven synthetic arrivals are scheduled. At zero buffer, all eleven enter the airborne queue. At seventy five, six point eight seven five are held and four point one two five are released. At one hundred fifty, the cap holds nine point three five and releases one point six five. This fraction rule is an assumption, not a per-flight seventy five minute timer. Another authored rule cancels flights whenever the ground queue exceeds thirty five, subtracting rounded eight percent of that queue. A synthetic queue of forty therefore cancels three and leaves thirty seven. Despite a comment mentioning three hour delays, that branch checks queue size, not each flight’s waiting time. Inspect the capacity, buffer and fuel assumptions, scrub the timeline, compare the synthetic manifest, and export the CSV. These outputs describe the page model rather than an operational dispatch plan.