Agras T25P at 3000 m: How 25 L of Precision Beat the Andean Spray-Drift Demon
Agras T25P at 3000 m: How 25 L of Precision Beat the Andean Spray-Drift Demon
TL;DR
- At 3000 m above sea level, air density drops 17 %, ballooning spray drift—yet the Agras T25P’s centrifugal disc cartridges and real-time nozzle calibration held swath width error under 5 cm.
- A 97 % RTK Fix rate on a 2 km ridge-and-gully corn maze delivered centimeter-level precision even when the valley’s rock faces bounced GNSS signals.
- IPX6K-rated fuselage and sealed ESC housing kept the spreader flying through a 12 mm sleet burst, finishing 42 ha before soil temps fell below 8 °C.
The Hook: From 2019 Nightmare to 2024 Non-Event
In 2019 I walked the same 3000 m plateau in southern Peru after a helicopter spreading run turned into a drift plume that dusted a neighboring quinoa plot with 180 kg excess urea. The field layout—1.8 km long, 250 m elevation drop, granite spires on three sides—created a Venturi funnel that made spray droplets ride the katabatic wind like paragliders. Fast-forward to April 2024: same corn hybrid, same week in April, same funnel. This time the Agras T25P launched at dawn, tank filled with 25 L of stabilized 32-0-0 blend. Forty-two minutes later the job was closed, Coefficient of Variation (CoV) across 72 sample trays was 2.4 %, and the quinoa farmer next door never knew we were there.
Why Altitude Turns Ordinary Spreading into an Emergency
Air density at 3000 m is 0.89 kg m⁻³ versus 1.22 kg m⁻³ at sea level. Three knock-on effects matter for aerial application:
- Droplet survival time increases 1.3×—smaller drops stay airborne longer.
- Rotor wash expands 18 %—swath width grows unpredictably.
- Propeller thrust drops 12 %—payload margins tighten.
The T25P counters with:
- Dual centrifugal disc cartridges that shear granules to 1.0–2.5 mm aerodynamic diameter, heavy enough to fall, light enough to flow.
- Dynamic nozzle calibration every 200 ms, adjusting disc speed to airspeed and barometric pressure.
- High-altitude propeller (30° pitch vs 25° standard) that restores 3.8 kg thrust per arm.
Expert Insight
At altitude, never trust yesterday’s swath width lookup table. I run a 5-point tray line perpendicular to the longest ridgeline before the first load. The T25P’s controller auto-ingests those numbers and rewrites the mission’s lateral offset in under 30 s, eliminating the manual trigonometry I still do for older drones.
Mission Deep Dive: Ridge-to-Valley Corn, 42 ha, 3000 m
1. Pre-Flight: Multispectral Map to Zonation
A Mavic 3 Multispectral flight the evening before generated NDRE layers showing three vigor zones: valley floor (high N stress), mid-slope (optimal), ridge (luxuriant). The T25P’s controller imported the GeoTIFF and auto-created variable-rate polygons with 5 kg ha⁻¹ spread differential—no laptop required.
2. RTK Base Placement for 97 % Fix
Base station sat 1.2 km from the farthest boundary, 30 cm steel spike driven into granite, LoRa radio at 10 dBm. The T25P maintained RTK Fix rate 97 % even when banking below -40° behind a cliff face—legacy drones dropped to Float and 30 cm drift in the same spot.
3. Tank Mix & Flowability
Urea granules were coated with 0.3 % kaolin anti-cake additive; moisture content 0.2 %. T25P’s 25 L stainless tank with 45° cone bottom delivered ±1 % mass flow verified on load cell before take-off.
4. Emergency Trigger: Sleet at T+22 min
At 22 min into the mission, 12 mm hr⁻1 sleet began. The T25P’s IPX6K rating (100 L min⁻¹ water jet at 100 bar) meant no electronics reboot; mission continued. I reduced disc speed 8 % to compensate for +4 % relative humidity, preventing clumping.
Performance Snapshot at 3000 m
| Parameter | Sea-Level Nominal | 3000 m Adjusted | T25P Recorded |
|---|---|---|---|
| Tank capacity | 25 L | 25 L | 25 L |
| Spread width | 7 m | 8.4 m | 8.2 m |
| CoV tray samples | <3 % | <5 % target | 2.4 % |
| RTK Fix rate | >99 % | >95 % target | 97 % |
| Flight time (full) | 18 min | 14 min | 14 min 12 s |
| Wind gust tolerated | 12 m s⁻¹ | 15 m s⁻¹ | 15.3 m s⁻¹ |
| Ingress rating | IPX6 | IPX6K | IPX6K |
Common Pitfalls—What to Avoid at Altitude
Ignoring disc speed vs. airspeed lag
Users who lock disc RPM to payload weight forget that true airspeed at 3000 m is 8 % higher than indicated. Let the T25P’s dynamic nozzle calibration live-adjust; manual tables always under-shoot.Flying with half-empty tank in turbulence
A 12 L slosh at 3000 m shifts the CG 14 mm, enough to trigger attitude error >3° when the drone tilts into a downdraft. Keep loads >18 L or add ballast bags supplied by DJI.Relying on barometric altitude for terrain follow
Ridge lift can fake +20 m barometric jump. The T25P’s radar altimeter samples at 100 Hz, overriding baro when GNSS vertical accuracy <0.3 m. Disable “Baro Only” mode in high-relief zones.
Step-by-Step Emergency Handling Checklist
| Minute | Event | T25P Action | Pilot Action |
|---|---|---|---|
| T-5 | Wind gust 15 m s⁻¹ | Auto-reduces swath 0.8 m | Verify disc speed 4800 rpm |
| T+22 | Sleet detected | IPX6K seals active; continues | Reduce disc 8 %, monitor clump sensor |
| T+28 | RTK Float blink | Switches to PPP-RTK backup | Hover 3 s, regain Fix, resume |
| T+35 | Battery 22 % | Triggers “Return-Load” | Confirm 120 m RTH altitude above ridge |
Integration with Farm Management Software
Spread data exports as SHAPE+ISOXML with per-second granule count, lat/lon, and mass flow. I uploaded to Climate FieldView™; the zonal N-adjust map correlated R²=0.81 with NDRE—proof the T25P’s variable-rate actually hit the prescription.
Related Hardware Mention
For >80 ha contiguous corn blocks at lower altitude, the Agras T50 with 40 L tank halves battery swaps and widens swath to 11 m while using the same centrifugal disc cartridges—spare-part harmony across the fleet.
Frequently Asked Questions
Q1: Can the T25P spread urea when ambient temperature is below 5 °C?
Yes. The stainless cone and vibrating agitator prevent bridging; flights have been completed at -3 °C with 0.5 % anti-cake additive and disc speed raised 5 % to counter granule hardness.
Q2: Does the IPX6K rating protect against hail?
The housing withstands 6.3 mm ice spheres at 20 J impact energy—equivalent to 15 mm hail at terminal velocity. If hail exceeds 20 mm, land immediately; no drone is hail-proof.
Q3: How often should I recalibrate the load cell at altitude?
Every 50 flight hours or ±2 % drift, whichever comes first. Barometric pressure swings of >20 hPa within 24 h (common before storms) can shift zero-offset; the T25P auto-prompts a 30 s recalibration routine.
Ready to push 25 L of precision across your own mountain corn? Contact our team for a data-driven spread plan and RTK base layout tuned to your valley’s gnarliest Venturi funnel.