How to avoid drift during drone spraying

Drone spraying is now becoming increasingly common across the globe. With the equipment and technique still in its infancy, however, there are mounting concerns about high levels of drift and questions regarding the quality of the application.

Dr. Dan Martin, a Research Engineer at the US Department of Agriculture’s Aerial Application Technology Research Unit in Texas quickly saw the enormous potential drones offer agriculture. But he also had his reservations.

Uniformity of sprayer deposition and application efficacy

As a specialist researcher in manned aerial applications, Dan realised while UAVs are increasingly being used for pesticide applications there was, and still is, a fair amount of uncertainty over their efficiency in terms of uniformity of sprayer deposition and application efficacy.

To find out more and how best to address these issues, he set up a research project with colleague, Dr. Wayne Woldt, from the Department of Biological Systems Engineering from the University of Nebraska. This practical, but also scientifically-based, study looks closely at the different elements that influence spray applications.

_{Text continues underneath image}

Tests at the USDA’s facility in Texas look at how any differences, such as nozzle size and position as well as operating height and speed impact the spray quality. – Photos: Dan Martin, USDA Aerial Application Technology Research Unit

Mitigate and avoid drift

“Our starting point was to see what equipment comes with on an ‘off the shelf’ drone,” he says. “The aim was to see how any differences, such as nozzle size and position as well as operating height and speed impact the spray quality. Also, importantly, we wanted to know what can be done to mitigate and avoid drift.”

The objectives of the research were to determine the ‘effect of application height and ground speed on the swath width (actual working width), spray pattern uniformity and droplet spectrum from different unmanned aerial application systems (UAAS)’.

The study looked closely at:

• Standard equipment on drones
• Calibration
• Nozzle choice
• Operating height
• Operating speed
• Swath width

Standard spraying drones

Wayne Woldt supplied his own DJI Agras MG 1. This popular 10 litre capacity drone is fairly typical of those in use in many areas. The researchers also turned to Florida-based, HSE-UAV, which supplied four spray drones of different sizes and capacities, which represent those mostly commonly used in the USA.

This provided an array of drones with six to eight rotors, with capacities from 10 litres to 20 litres and four to six nozzles – either fixed to booms or mounted beneath the rotors.

HSE-UAV says it participated in the trials to help its customers understand how to make effective applications, as well learn how it can improve its hardware and spray systems. It is also looking to develop an industry-wide standard.

“What we found, straight away, is when it comes to spray equipment they are fairly basic – with not one of them fitted with a pressure gauge. Also nearly all of them came equipped with 005 to 0067 sized nozzles, which create fine droplets,” explains Dan.

_{Text continues underneath image}

During the static calibration procedure the researchers also assessed set-up and level of equipment that comes with a standard ‘off the shelf’ spraying drone.

Calibration is crucial

Dan started off by calibrating the sprayers, not an easy task when none are fitted with pressure gauges. So he made an adaptor and quick coupler to fit into the drone’s plumbing system. “The set pressure we discovered varied from 33psi (2.27 bar) to 72psi (5 bar),” he adds.

“While it’s important to reduce weight on a drone, it’s also crucial it’s calibrated correctly. For all sprayers to do that you need to know the pressure and forward speed and without a gauge you are missing a vital part of the equation. What it needs is a small transducer that displays pressure on the operator terminal (or ground station),” he suggests.

Dan and the team set the appropriate pressure and measured the flowrate by collecting the output from each nozzle in cups over a set time. This is a well known procedure for operators of conventional boom sprayers.

_{Text continues underneath images}

Application rate of 2gal/acre

With manned aerial applications common in the USA, there is a wide range of approved products, which have to be applied according to the label recommendations. For most this is an application rate of 2gal/acre, equivalent to about 20 litre/ha, which was also used for the experiments.

You can dramatically cut drift by simply selecting the right nozzle

It’s important to know the drone pumps outputs. We found this varied among the different models with the smallest, V6A, unable to deliver 20 litre/ha even at the lowest speed with the 005 nozzle. HSE-UAV has since ‘retired’ this model from its offering.

Forward speed is also an important calibration factor, and for the research the team tested each at four ground speeds – 1m/s, 3m/s, 5m/s and 7m/s.

Actual working width

Lastly, operators also need to know the actual working width to determine the area. “Swath width is determined not only by the nozzle spacing, but also the type of aircraft. It could also be effected by the working height as well as speed. So to test the widths we carried out applications at three heights – 2m, 3m and 4m,” he explains.

Slower speeds had little effect on swath but greatly impacted application rate. Application heights of 2m and 3m allowed for the spray pattern to mature while still mitigating drift by staying lower to the ground. Tests indicated that effective swath was most effected by nozzle (fine, medium, coarse) and size of aircraft.

Test procedure

Tests were carried out at the three application heights, at the four ground speeds and with four repetitions. Each was also tested with three different nozzles, to check the swath width, coverage and drift and assess the different droplet spectra.

These included:

• Fine: XR 110 01 – ~200-250 microns
• Medium: TeeJet TT 110 01 – ~300-350 microns @ 2.0 to 3.5 bar
• Medium to coarse: Lechler IDK 110 01 – ~350-400 microns @ 3.0-6.0 bar

All applied water with liquid fluorescent dye, plus a surfactant, at 0.25% by volume, to replicate the characteristics of a plant protection product. Droplets were collected on water sensitive paper, across the working width, to assess the drop size, number and coverage to characterise the droplet spectrum. The results were analysed using the DropletScan system.

_{Text continues underneath video}

Dramatically cutting drift

“With 20 repetitions and averaging the results, the results were very clear: you can dramatically cut drift by simply selecting the right nozzle. This confirmed our initial thought that we were expecting to see less drift from the medium and coarse nozzles,” says Dan.

It also emphasises how nozzle selection is hugely important as a way to reduce driftable fines, which he observed are pushed down and outwards by the rotors on UAVs. “I measured a windspeed of 15.5m/s (35mph) under the rotors of my spray drone hovering at 3m. While this will effect drift, one benefit we are now researching is how this may assist crop penetration, helping droplets to move into the canopy and under the leaves,” he adds.

Many drones are coming equipped with 80 005 nozzles, which create huge amounts of drift, which is obvious from many of the images of spraying operations – often in Asia.

“HSE-UAV, which supplied its drones for our experiments fitted with Lechler 80 005s, has already changed to offering a choice of Lechler IDK 110 01, 015 or O2 as standard following the tests,” adds Dan. “The 015s work well as do even an 02 in areas where drift could be a problem.”

_{Text continues underneath images}

Swath width variation

Whether nozzles are mounted beneath the rotors or on a boom, the research found position doesn’t necessary set the swath width on a UAV. Switching from fine to medium nozzles, however, will narrow the swath. “We haven’t been able to test effect of nozzle location on drift yet. Probably next year. But our best guess is that it will not affect drift nearly as much as nozzle selection,” he adds.

Each of the spraying drones’ swath width was measured with the same technique as that used for manned aerial applications. This involves applying the product over a cotton string and then analysing this in the lab to determine the quantities deposited along its length.

All the flights were carried out at 3m, which the researchers had already determined to be the optimum application height.

“We found this to be the ‘sweet spot’ that allows the pattern to fully develop. UAVs need to be operated in Visual Line of Sight (VLOS) and at 3m its possible to keep it in view, even over undulating land, without it disappearing into dips. It is also above head height and helps avoid any obstacles.”

The team initially used a 10m length of string, but extended this to 15.25m to catch the end of the patterns and take account of the wind.

Reduce the risk of drift

“At a height of 3m we discovered it provides the optimum swath width and coverage with the selected nozzles. The spray has plenty of time under the rotors and this helps to reduce the risk of drift,” explains Dan.

The 20 litre M8A PRO drone, which has its nozzles fitted to a boom, initially produced a noticeable ‘spike’ in the pattern in the middle. This was overcome simply by adjusting the spacing between the in-board nozzles.

Measured swath widths varied between models from 3.6m to 9m and, with fine nozzles, this ranged from 7.5m and 9.0m.

Watch your speed

Operating speed is limited by the drone’s capabilities. But it is an important consideration when calibrating for flowrate and can also influence the application rate and coverage.

Researchers operated the UAVS at four ground speeds – 1m/s, 3m/s, 5m/s and 7m/s, calibrating the sprayers accordingly.

Speed and height actually showed little effect on effective swath. Most important parameters were nozzle and size of aircraft (power). Although 7m/s did provide the largest effective swath for the V6A, it is not statistically different from the other speeds. While 3m/s was the best speed for the MG-1, again it is not statistically different from the other speeds. Higher speed yields increase productivity, but application rates on the label MUST be met.

_{Text continues underneath images}

Popularity of drones is set to soar globally as countries grant operators permission to apply crop protection products. Read why drone spraying takes off as regulations relax worldwide.