By Wesley Alexander • September 14, 2026 • 8 min read

On Friday, September 11, the FAA quietly posted a document that describes more daily drone traffic than anything the agency has ever put out for public review. The Draft Environmental Assessment for Zipline's proposed operations in multiple Texas metropolitan areas covers Houston, San Antonio, Austin, Amarillo, and El Paso, and the ceiling numbers are worth reading twice: up to 220 charging towers, 4,400 pickup kiosks, 220,000 deliveries per day, and because each delivery involves three flight legs, as many as 660,000 aircraft transits per day across the five metros. Around the clock, seven days a week, at 330 feet AGL and 70 mph.

The comment window is open for 30 days from September 11. If history is a guide, almost nobody will use it. When the FAA ran the same process for Zipline's Dallas-Fort Worth expansion last year, the final EA records that the agency received exactly one public comment before signing a Finding of No Significant Impact that December.

One comment, for an operation that authorized up to 30,000 flights a day across 10,904 square miles of North Texas. This filing is roughly seven times larger. That asymmetry between the scale of what is being decided and the attention it gets is the story, and it cuts differently depending on which side of the industry you sit on.

What the filing actually asks for

Strip away the NEPA structure and the mechanics are familiar. Zipline holds a standard Part 135 air carrier certificate and a Section 44807 exemption that together let it carry property for hire beyond visual line of sight with its Platform 2 aircraft. What actually turns that authority on in a given city is the Operations Specifications, the OpSpecs, which name the geographic areas where the certificate holder may operate. Amending OpSpecs to add five metro areas is a major federal action under NEPA, which is why this environmental assessment exists and why the public gets a say at all.

The proposed architecture, per the Draft EA:

If Zipline ever wants to exceed those allocations, the document is explicit that a new NEPA review is required. The numbers in this EA are the envelope, not the forecast; DFW's own phasing plan started at five sites and 20 flights a day. But envelopes are what get approved, and this is the envelope the FAA is being asked to sign.

The bottleneck is not Part 108, and this proves it

There is a persistent belief in this industry that scaled drone delivery is waiting on the Part 108 BVLOS rule. This filing is a useful corrective. Zipline is not waiting for Part 108. Neither is Amazon, whose Prime Air Texas expansion cleared its own FONSI in February. The scaling machinery that exists today is Part 135 plus a 44807 exemption plus OpSpecs amendments, and the pacing item in that machinery is environmental review: study areas, endangered-species consultation, Section 106 historic-property findings, noise modeling, tribal consultation across fourteen nations, and a public comment period.

For operators building delivery or BVLOS programs, that has two practical implications. First, your regulatory roadmap should treat NEPA as a schedule line item, not a formality. DFW went from draft EA to FONSI in about six months; this five-metro filing is bigger and touches more consultation tracks, and the FAA must respond to every substantive comment in the final EA. Second, these documents are the closest thing you will get to a published playbook for how the FAA thinks about siting scaled UAS infrastructure. The Texas EA specifies noise standoffs of 160 to 560 feet between infrastructure and noise-sensitive properties depending on airspace classification, keep-out zones around airports and open-air assemblies, and per-point overflight caps. If you are drafting a ConOps for anything that scales, those constraints preview the questions you will eventually be asked. We saw the same dynamic in Amazon's Baton Rouge noise filing: the environmental paperwork is where the real operating limits get written.

The airspace question the EA does not answer for you

Here is the part that should interest every operator who is not Zipline. The 330-foot cruise altitude sits squarely inside the band where Part 107 operations live under the 400-foot ceiling, where DFR programs launch and recover, where helicopter EMS and law enforcement work low, and where agricultural and inspection flights spend their days. An environmental assessment evaluates noise, wildlife, and historic properties. It is not the venue that deconflicts your survey flight from a delivery corridor running 195,000 transits a day across metro Houston.

That deconfliction burden lands on operating rules, Remote ID, strategic separation in Zipline's OpSpecs, and, frankly, on you. A test pilot's read of these numbers: at this density, the low-altitude environment in these five metros stops being empty by default. Preflight planning that assumes nothing else is out there at 300 feet was already lazy; in a city with authorized delivery corridors it becomes a genuine hazard assumption. Know where the operating areas are, treat published delivery infrastructure like you treat a heliport, and log your own position and altitude discipline accordingly. If your program depends on predictable access to that band, the comment window is your one formal opportunity to put airspace-integration concerns into a record the FAA is obligated to answer. Comments go to 9-FAA-Drone-Environmental@faa.gov within the 30-day window that opened September 11.

It is also worth saying plainly: scale changes the safety math. Zipline's fleet has an enormous safety record over Rwanda, Ghana, Arkansas, and Texas, and the highway parachute event we analyzed in September showed the failure chain working as designed. But 660,000 daily transits is a different exposure regime than anything currently flying, and contingency events that are vanishingly rare per flight become weekly occurrences at fleet scale. The EA does not analyze that; NEPA is not a safety case. The OpSpecs review is, and it happens without public input.

What to do with this if you are building toward scale

If you run or are planning a delivery, DFR, or infrastructure BVLOS program, spend an hour with this document and the DFW final EA side by side. Extract the siting standoffs, the phasing tables, the keep-out logic, and the consultation tracks, and map them onto your own concept of operations before a regulator does it for you. Approval is a paperwork architecture, and the operators who scale are the ones who build the readiness file before they need it. The UAVHQ BVLOS Playbook includes ConOps and planning templates structured around exactly these questions, and the weekly UAVHQ dispatch on Substack will track this docket through the final EA and whatever the FAA decides.

The last time the agency asked Texas what it thought about tens of thousands of daily drone flights, one person answered. This time the ask is seven times bigger. Whether the record stays that quiet is up to the people who actually use this airspace.

Sources

  1. FAA, "Draft Environmental Assessment for Zipline International Inc. Proposed Drone Package Delivery Operations in Multiple Texas Metropolitan Areas," September 2026
  2. FAA, "Public Involvement and Environmental Review for Drone Operations" (NEPA and drones portal)
  3. FAA, "Final Environmental Assessment for Zipline International Inc. Proposed Drone Package Delivery Operations in Dallas–Fort Worth, Texas," December 2025
  4. FAA, "Finding of No Significant Impact and Record of Decision, Zipline Dallas–Fort Worth," December 23, 2025