Air-Assist Booms

Posted on by Jason Deveau (Spray_Guy)

About Jason Deveau (Spray_Guy)

Dr. Jason Deveau (@spray_guy) has been the OMAFRA Application Technology Specialist since '08. He researches and teaches methods to improve the safe, effective and efficient application of agricultural inputs in specialty crops, field crops and controlled environments. He is the co-administrator of Sprayers101, co-author of the Airblast101 Textbook, a slow cyclist and a slower runner.

See all posts by Jason Deveau (Spray_Guy).

Air-assisted boom sprayers have been around since the 80s (likely longer). They’re prevalent in Europe but are not common in North America. Some vegetable and berry growers use them, but few if any field croppers. This is most likely because most air-assisted sprayers are trailed. Trailed sprayers don’t have the high clearance or boom stability to tolerate the high travel speeds preferred by North American field croppers. Or, perhaps, operators feel the advantages of air-assist don’t outweigh the price-tag associated with the feature.

Hardi Commander (118 foot boom) with TWIN air-assist
Hardi Commander (118 foot boom) with TWIN air-assist

Given my background with horticultural airblast sprayers, I’m a proponent for using air when spraying. It opens the crop canopy, exposes all foliar surfaces, carries small droplets to the target (improving coverage and reducing drift) and it extends the spray window by out-competing moderate winds. Some manufacturers of air-assist for horizontal booms claim that it improves coverage so much that less spray mix is required per hectare.

With benefits like that, why don’t we see more of these in North America? What does Europe know that we don’t? To find out, we ran a demonstration at Canada’s Outdoor Farm Show. We used a Hardi Commander with TWIN air-assistance to spray water into soybeans planted on seven inch centres (a moderately dense broadleaf canopy).

The demo treatments

The sprayer was calibrated to apply water 93.5 L/ha (10 gpa) at 2.75 bar (40 psi) at 9.7 km/h (6 mph). This was dialled into the rate controller and intended to represent a typical fungicide application. The boom was suspended 50 cm (20 inches) above the top of the canopy. On one side of the boom, we ran yellow mini drift nozzles (MD 11002’s) to create a coarse spray quality, and on the other side we ran conventional yellow flat fans (F 11002’s) to produce a fine spray quality.

Water-sensitive paper was attached to rods that were placed in the soybeans. The papers were placed at three canopy depths: at the top, midway down and at the bottom of the canopy. Papers were oriented both face-up and face-down. Following each application, papers were collected for digital analysis using “DepositScan” which determines the percent of the paper covered with spray, and the droplet density. Both of these factors contribute to overall coverage.

We collected papers for three treatments:

  1. Fine spray quality, No air assist
  2. Coarse spray quality, No air assist
  3. Fine spray quality, Air assist
Figure 2 - Water-sensitive papers were placed at three levels in a dense soybean canopy, facing up and down, for three conditions. Condition 1 - Air off, conventional 11002’s (medium-fine spray quality). Condition 2 - Air off, mini drift AI11002’s (very coarse spray quality). Condition 3 - Air on, conventional 11002’s (medium-fine spray quality).
Figure 2 – Water-sensitive papers were placed at three levels in a dense soybean canopy, facing up and down, for three treatments. Treatment 1 (Fine spray quality, No air assist). Treatment 2 (Coarse spray quality, No air assist). Treatment 3 (Fine spray quality, Air assist).

We held two demos per day at noon and 3:00 pm for three days, giving us six sets of papers to analyse for each treatment. The weather spanned from 25-29°C, 30-58% relative humidity and winds of variable direction from 3-11 km/h.

This was a very simple randomized complete block design. A more fulsome experiment would require many, many more passes under more stable conditions. For example, we set the angle of the air and nozzles to about 30° forward and the air speed at maximum, which wasn’t necessarily correct. Ideally, these settings should have been fine-tuned to match the forward speed of the sprayer, the density of the crop and the weather conditions. Additionally, droplets under 50 microns do not register in the DepositScan software, even though they can still be discerned with the eye (only just). This is a disconnect between coverage and potential efficacy that may or may not make a significant difference.

And so, caveats aside, the following graph illustrates the mean percent coverage and mean droplet density for papers in each treatment, for papers that were facing up (figure 3). Standard error of the mean is presented alongside the average (x% ± y).

Results

Figure 3 – Average percent coverage (red) and droplet density (blue) for upward-facing water-sensitive papers in three canopy depths for each of three conditions. Averages rounded to the nearest 0.5 and Standard Error is indicated. * indicates significance with 95% confidence.
Figure 3 – Average percent coverage (red) and droplet density (blue) for upward-facing water-sensitive papers in three canopy depths for each of three treatments. Averages rounded to the nearest 0.5 +/- standard error. “*” indicates significance with 95% confidence.

Treatment 1 (Fine, No Air) reflects a typical coverage pattern for a dense canopy. Coverage declines as a function of canopy depth because spray droplets are intercepted by plant material before they reach the ground. This is particularly evident with broadleaf canopies that create shading. It should be noted that there was an obvious plume of spray drift during these applications (see Figure 4).

Figure 4 – The effect of air-assist on downwind drift from a medium-fine spray quality. Note that the nozzles and air are directed 30° forward. When sprayed over bare ground, the air-assist bounces spray back up, as pictured here. However, when sprayed into a canopy with the correct air settings, bounce (and drift) is virtually eliminated.
Figure 4 – The effect of air-assist on downwind drift from a medium-fine spray quality. Note that the nozzles and air are directed 30° forward. When sprayed over bare ground, the air-assist bounces spray back up, as pictured here. However, when sprayed into a canopy with the correct air settings, bounce (and drift) is virtually eliminated.

Treatment 2 (Coarse, No Air) follows the same coverage trend as Treatment 1. This treatment represents much larger, and fewer, droplets than Treatment 1, and yet the only obvious difference is reduced coverage in the middle of the canopy. There was little or no plume of spray drift during these applications.

Treatment 3 (Fine, Air) also followed the trend of reduced coverage as a function of canopy depth. Mean coverage was higher at the top of the canopy compared to the other two treatments. In fact, according to an ANOVA, droplet density was significantly higher in this canopy position than the other treatments, with 95% confidence. While mean coverage in the middle of the canopy was more than 2x that of Treatment 2 (see Figure 5), it was not statistically significant. There was no apparent difference at the bottom of the canopy. Unlike Treatment 1, there was little or no plume of spray drift during these applications (see Figure 4).

Figure 5 – Upward-facing water-sensitive paper from mid-way into the canopy (position B) for condition 2 (very coarse droplets, air off) and condition 3 (medium-fine droplets, air on). The difference in coverage is obvious.
Figure 5 – Upward-facing water-sensitive paper from mid-way into the canopy (position B) for Treatment 2 (Coarse spray quality, no air assist) and Treatment 3 (fine spray quality, Air assist). The difference in coverage is obvious.

DepositScan was unable to detect coverage on any of the downward-facing papers. However, close visual inspection did reveal differences. Perhaps unsurprisingly, Treatment 2 (Coarse, No air)  did not produce any coverage. Large droplets do not change direction mid-flight unless acted upon by some other force. Droplets can bounce and shatter, but that did not occur here. Conversely, the medium-fine droplets created in Treatment 1 (Fine, No Air) and Treatment 3 (Fine, Air) did leave trace coverage on the downward-facing papers. Generally no more than 10-30 droplets on the entire 1×3 inch paper, representing less than 1% total surface coverage. It could not be determined if the air used in Treatment 3 improved coverage over that of Treatment 1.

So, did it work?

Many peer-reviewed journal publications say yes. A perfunctory literature review reveals improved coverage in the middle and lower portions of cotton, potato, soybean and wheat canopies. Some of these experiments were based on coverage using fluorescent dyes and/or water-sensitive paper. Others were based on efficacy to determine if pest control was improved. Some recommended angling the air and nozzles forward 20-25°, some proposed 30° backwards and most claimed that the higher the air speed, the better the penetration and lower canopy coverage at the expense of coverage in the higher canopy.

As for our results, some interpretation is required.

If an operator is spraying a contact product with limited or no translocative properties, then coverage becomes very important. In order to improve coverage, higher volumes and finer droplets combined with slower travel speeds are often advised. However, operators prefer to use less water and drive faster, and fine droplets contribute to spray drift.

When we used medium-fine droplets with no air assist, coverage was very good (Figure 3) and better than coverage obtained using very coarse droplets. However, spray drift was unacceptable (Figure 4). When air-assist was engaged, mean spray coverage was improved (Figure 3), and drift was greatly reduced. An important question springs to mind:

If the drift was blown back into the crop, why wasn’t coverage more significantly improved? The droplets had to go somewhere…

While not always statistically significant, mean coverage was improved in the top and middle of the canopy. My hypothesis is that we didn’t set the air-assist as carefully as we should have and as a result, droplets were not carried deeply enough into the canopy. Further, the angled spray likely improved stem coverage in the same way that angled nozzles improve coverage on any vertical target, but we didn’t explore that possibility. Finally, we must reiterate the resolution limitation of water-sensitive paper – tiny drops don’t register in the digital analysis, but that doesn’t mean they weren’t there and providing chemical control.

Final thoughts

In the end, there are obvious advantages to the air-assist mechanism. The ability to employ a finer spray quality when required, while greatly reducing spray drift and combating inclement weather to extend the spray window are all appealing features. Research has clearly demonstrated that deep-canopy spray coverage and overall efficacy are improved when this system is properly adjusted to match spray conditions. Given my limited experience with the system, I’m not comfortable with suggesting it warrants lower dosages because of the expertise required to adjust the system. However, experienced operators have accomplished it in Europe.

As new high-clearance, self-propelled sprayers (such as the Hardi Alpha) are introduced to North America, I think we’ll be seeing more air-assist options on boom sprayers.

You might also like: