Tag: evaporation

  • The Droplet Size Debate

    The Droplet Size Debate

    Funny how some issues never go away. For as long as I’ve been in the sprayer business, the question of ideal droplet size for pesticide application has remained a hot topic.  At its root are the basic facts that small droplets provide better coverage, making better use of water, but large droplets drift less.  So why are we still debating this? Because we need both of these properties to be efficient, effective, and environmentally responsible. Ultimately, the droplet size question is reduced into one of values, where everyone’s individual priorities play a role. 

    First, let’s talk about basic principles. To be effective, an active ingredient must make its way from the nozzle to the site of action in the target organism. On the way, it encounters several obstacles as summarized by Brian Young in 1986.

    Figure 1: The dose transfer process of pesticides (after Young, 1986)

    After atomization and before impaction, the spray encounters two main losses, evaporation and drift. Both of these are more severe for smaller droplets. Smaller droplets have a greater ratio of surface area to volume for any given spray volume, and can evaporate to a much smaller size, even to dryness depending on the formulation, in seconds. For water-soluble formulations, one consequence is lower uptake. Oily formulations may maintain efficacy, but neither type can escape the second effect, spray drift.

    Figure 2: Time to evaporate all water from droplets of various sizes, based on the “two-fluid” model developed by Wanner (1980). Based on 0.8% v/v non-volatile, non-soluble addition, 20 ºC, and 50% RH. This model suggests that final droplet diameter is 20% of initial diameter. Reproduced from Microclimate and Spray Dispersion by Bache and Johnstone (1992, Ellis Horwood Ltd).

    Small droplets are more susceptible to displacement by wind currents due to their small mass. There is no magical size above which drift is no longer possible, but we’ve generally used diameters of 100, 150, or 200 µm as a theoretical cutoff. The proportion of the spray’s volume in droplets smaller than these diameters can be called “drift potential”, and this value is useful to measure the impact of nozzle type, pressure, or formulation on that phenomenon.

    But it’s not quite that simple. Even a small droplet may resist drift if its exposure to wind is limited, perhaps through a protective shield shroud, or lower boom height. Or by increasing its speed through air assist. Higher energy droplets resist displacement.

    These mitigating strategies aren’t lost on sprayer manufacturers who have used them for decades to build lower drift sprayers.

    The next phase of the dose transfer process is interception. The droplet has to encounter its target, but the process is mostly coincidence. Simply put, the target has to be in the way of the droplet’s flight path for the two to meet. Denser canopies are therefore more effectively targeted. A larger number of droplets (smaller droplets or more carrier) also improve the odds. But it’s not that simple. Flight paths can change. That’s where small droplets are more inventive. Because they respond to small air currents, and because such small currents surround most objects, the smaller droplets can weave around objects, following the small eddies generated by air flows. As a result, we’re more likely to find smaller droplets further down in denser, more complex canopies where the eye can’t follow. They simply cascade through.

    Larger droplets, on the other hand, resist displacement by air and travel in straighter lines. They tend to hit the objects they encounter. For that reason, larger droplets are intercepted by the first object they reach and only make their way deeper into a canopy if the path is clear. In other words, vertical, sparser objects allow larger droplets to pass by.

    These properties are related to the droplet’s inertia, and are best described by a parameter known as “stop distance”. Assuming an initial velocity, stop distance is the distance required by a droplet to slow to its terminal velocity.

    Figure 3: Stop distance as a function of droplet size. Assuming a 20 m/s initial velocity (similar to exit velocity of a hydraulic nozzle) and gravity assistance. Note that smaller droplets without the benefit of air assist lose their initial velocity within a few cm of the nozzle exit. Reproduced from Microclimate and Spray Dispersion by Bache and Johnstone (1992, Ellis Horwood Ltd).

    These characteristics, combined with the aerodynamic properties of objects such as tiny insects, cotyledons, leaves, stems, etc. govern the collection efficiency of sprays. Small, slow moving droplets are thus best captured by small objects that don’t create strong enough deflections of airflow to steer the droplets past. Large objects that redirect air around them very effectively are better collectors of the larger or faster droplets whose kinetic energy can guide them through this turbulence. It’s also a matter of probability, as the smaller objects tend to have a lower likelihood of encountering the relatively scarce large droplets of any given spray.

    But once again, that’s not the end of the story. Interception is followed by a critical stage, retention. Objects must be able to hold onto the droplets they intercept. Slow motion video has shown that droplets flatten out on contact with an object as the liquid converts impaction velocity into lateral spread. Once at full extension, the flattened droplets will collapse even beyond their original round shape, pushing them away from the surface and possibly causing rebound. A rebounding droplet may eventually land on target, but that would be a matter of fortune. It’s better if the leaf can offer enough adhesion, diminishing the power of the rebound oscillation, allowing droplet to stick the first time.

    Figure 4: Droplet deformation during impact (C. Hao, et al. 2015. Superhydrophobic-like tunable droplet bouncing on slippery liquid interfaces. Nature Communications. August 2015).

    Small droplets have less mass, and tend to be retained more easily. But more than size is at play here. The morphology and chemistry of the leaf surface is also important, with crystalline or more oily surfaces offering less adhesion for droplets. The physico-chemical properties of the spray mixture becomes important, as characteristics such as dynamic surface tension and visco-elasticity affect spray retention. These properties are optimized through the product formulation effort, and possibly via adjuvants added to the tank.

    We sometimes classify targets as “easy to wet” or “difficult to wet” to summarize these properties. Most grassy plants (foxtails, cereals) are difficult to wet (there are exceptions, such as the sedges) and broadleaf plants vary from the easy to wet pigweeds to the difficult to wet lambsquarters and brassicas. Easy to wet species can retain larger droplets than difficult to wet species, and that’s one reason why finer sprays are preferred for grassy weed control (leaf orientation and size are another).

    Figure 5: Droplet deformation, and surfactant molecule alignment, during impaction. The inability of surfactants to reach optimal alignment quickly, and for the target surface to absorb these forces, leads to rebound.

    A few words about surface tension. Although surfactants reduce surface tension and facilitate spreading, this may not be enough to improve spray retention. To be effective, surfactant molecules need to align themselves with the surface of the droplet so they can be a “bridge” at the interface where the droplet meets the target surface. This takes time. The oscillations that occur during impaction continuously create new surfaces, and if surfactant molecules don’t follow suit immediately, the droplet will behave as if no surfactant is present.  Specialists measure “dynamic” surface tension, i.e., the surface tension at young surface ages – a few milliseconds – to better predict spray retention. Very young surface ages have surface tensions of plain water, even with a surfactant present. Only certain surfactants, or higher concentrations of surfactants, can actually improve spray retention.

    When air-induced nozzles were introduced in the mid 1990s, one of their claims was the improved spray retention due to air inclusions (bubbles) in the individual droplets. These bubbles made the droplets lighter, and also reduced their internal integrity, promoting breakup on impaction. As a result, the coarser sprays they produced actually had some of the same efficacy performance as the finer sprays they replaced. And indeed, research showed that coarser, air-induced sprays did in fact maintain good performance. Interestingly, performance of non-air-induced coarse sprays used with pulse-width modulation also showed similar robustness of performance. Research comparing air-induced to conventional sprays of similar droplet size rarely showed differences, and when they occurred, they were small in magnitude and could be corrected through improved pattern overlap.

    Figure 6: Air Bubbles in spray droplets (Source: EI Operator. Believed to originate with Silsoe Research Institute, UK)

    One reason larger droplets still work well is due to the pre-orifice designs of modern low-drift nozzles. This design reduces the internal pressure of the nozzle itself, with the effect being a slower moving large droplet. This reduced velocity takes away some of the force at impaction, reducing rebound.

    Figure 7: Droplet velocity of larger droplets is reduced by lower pressures from pre-orifice and air-induced design nozzles. Lower velocities reduce droplet rebound.

    Another neat effect of coarser sprays is their ability to entrain air. All sprays move air (simply spray into a bucket to see this), and larger droplets do this better and for longer distances. The entrained air is a form of air assist for the smaller droplets, increasing their average velocity and thus reducing their drift potential while they move in the spray pattern. 

    The final stage of the dose transfer process is deposit formation and biological effect, and that’s where we once again see differences attributable to droplet size.

    Once established on a target surface, the active ingredient usually needs to move to its site of action. In some cases, resting on the surface is sufficient, it depends on the specific product. But for the majority of herbicides, the active ingredient must move across the cuticle into the cytoplasm where it eventually migrates to the enzymes involved in photosynthesis or biosynthesis of fatty- or amino acids.  The cuticle is waxy, with only a few water-loving pathways and the uptake process is basically driven by diffusion and concentration gradients. As such, it is more effective when the product is in solution and the longer the droplet can stay wet, the better.  That’s one reason why spraying during hot, dry days may reduce performance. Again, it depends on the formulation and the mode of action. Too high a concentration can damage membranes, physiologically isolating the active ingredient and reducing its subsequent translocation. It’s always a balancing act.

    If you’ve been keeping track of the score, it’s more or less a tie between large and small droplets. One deposits better and makes more efficient use of lower water volumes, while the other has lower losses from drift and evaporation, helps smaller droplets resist drift, and may improve uptake of some products.

    And this draw is why the venerable hydraulic nozzle has been so successful for so many decades. Hydraulic atomization, by its nature, creates a wide diversity of droplet sizes, ranging from 5 to 2000 µm or greater. As Dr. Ralph Brown of the University of Guelph used to say, this nozzle provides a drop for all seasons. Some small ones for coverage and retention in hard to reach places, and some large ones for uptake and drift-reduction. The result is a robust delivery system that provides reliable results on many different targets under many conditions. In recognition of the heterogeneity of sprays, we don’t refer to specific droplet sizes, but rather their composite, grouped into international categories of Spray Quality such as Medium, Coarse, and Very Coarse.

    Our challenge is to find the spray quality sweet spot, the ideal blend of these contradictory and yet complementary features of our agricultural sprays. And I believe that task is very achievable. Simply put, broadcast agricultural sprays in field crops work reliably when applied as Coarse and Very Coarse sprays in volumes between 7 and 12 US gpa. There is no need to spray any finer than Coarse for good efficacy, as coverage is already sufficient and any additional coverage has small marginal returns. There is, however, value in adding more water when canopies are denser or when leaf area index grows as the crop matures. To gain coverage, adding water is preferred to reducing droplet size because of the value of environmental protection. It so happens that Coarse to Very Coarse sprays provide or ecxeeed the drift protection required by most agricultural labels.

    There is occasional reason for spraying even coarser than what I’ve suggested. It’s certainly required by law for dicamba products on Xtend traited soybeans and cotton, but even then, only in conjunction with higher water volumes to offset losses in droplet numbers.  In practice, moving to Extremely Coarse or Ultra Coarse sprays may allow an application to proceed in higher than average wind without adding drift risk. The use of some additional water is a relatively small price to pay for that additional capability.

    There will always be opportunities for efficacy improvement in specific cases for those willing to spend the extra time to optimize that situation. That’s one of the reasons I’m excited to see the widespread adoption of pulse width modulation (PWM) in the industry, allowing users to change spray pressure and therefore spray quality with no impact on application rate or travel speed. Or the introduction of nozzle switching from the cab, employing the optimal atomizer for a specific situation. Although it remains difficult to define the ideal spray, selecting a spray quality has never been so easy.

  • Don’t try this tempting shortcut

    Don’t try this tempting shortcut

    There’s a call that I’ve been getting for 20 years now. It came again this week. Someone has a twincap with two small air-induced tips, and they’re applying herbicides and fungicides with low water volumes, often 5 gpa, sometimes less. They call because they want to know how much wind they can spray in. Is 30 km/h OK? They want my blessing.

    I don’t need to hear much more. Some nozzles are sold entirely on the premise that they provide superior coverage – more droplets per square inch – and that this improved coverage permits the reduction of water volumes. Furthermore, the claim goes, when water is reduced, the spray concentration increases and the whole darn package just works a lot faster and better.

    This line of thinking is as old as spraying itself. Applicators seek pesticide performance as well as productivity, and this approach gives them both. The proponents are well aware of their customers’ desires, and sell into it. “Use these tips and cut back on water. Any more than this just runs off anyways. You’ll get better coverage and better performance, get more spraying done.” It’s a convincing argument. Get an edge on your neighbour, the person who’s not in on the secret and is wasting time and water.

    Why don’t I embrace it? There are a few reasons.

    First, it doesn’t tell the whole story. Invariably it involves a twin nozzle setup. Use two nozzles, get more droplets, right? If that were true, believe me, I’d be advocating for quintuples.

    Fact is that the only factors that change droplet numbers are droplet size (spray quality) and water volume. Want more droplets at the same water volume? Make the spray finer. Want to keep spray quality and add droplets? Add water (not nozzles).

    The easiest way to improve coverage at the same volume is to use a finer nozzle, or to increase spray pressure. Depending on how far you go, you could make the spray finer and cut water, and still have more droplets per square inch.

    The hardest way to improve coverage is to purchase a twincap and buy two nozzles, each of them half the size. True, within any given nozzle type, smaller sized tips usually generate finer sprays. But why bother with two tips? They’re more expensive and plug more.

    If someone asks me how to improve coverage without changing water volume, I usually tell them to speed up a few mph. The rate controller will increase pressure and the spray gets finer. If speeding up is not possible, get one size smaller nozzle and run at higher pressure, same speed. Or keep nozzle and speed, and add some gpa, pressure will go up. It’s that easy. No twins necessary.

    Second, the twin nozzle/low volume approach exaggerates the value of the twin nozzle for herbicides. With small plants and relatively open canopies in the early season, plus our high booms and travel speeds, the twin tips are not adding a lot, if anything at all, to coverage. It remains a sum of droplet size and water volume, the angle is not important at this stage. Deposit is by turbulence and wind, most of the time.

    Third, low volume believers ignore a few potential problems. Drift is a big one. Low volume, fine spray operators are surrounded by nervous neighbours. They have fewer hours per day during which drift is acceptably low. And they definitely should not be on the field when wind is at 30 km/h. Basically, they’re a bit uncomfortable (at least they should be) and get less done per day.

    Another potential problem is evaporation. Most sprays, even when applied at lower volumes, are still 90% or more water. The same volume of water evaporates much quicker when atomized into smaller droplets. This has two main downsides: On their way to the canopy, small droplets evaporate and become even more drift prone, and may not impact at all. Those that impact evaporate shortly thereafter. Research has shown that pesticide uptake is better from wet than dry deposits.

    When Delta T (dry bulb minus wet bulb temperature) is high, evaporation can be so strong that it reduces pesticide performance or causes solvent burn. Fine sprays make it worse.

    I also hear about the use of oily adjuvants to control evaporation from small droplets. This could be even more dangerous. Small droplets drift, and evaporation to dryness is actually helpful in reducing the impact of that drift. How? It makes the small droplets disappear, with their remnants dispersing into the turbulent atmosphere. With oily adjuvants, the small droplets stick around and stay potent and their drift damage is much worse.

    Lastly, the practice is possibly off label. Water volume and spray quality label statements are designed to offer good performance and acceptable drift risk. While that part of the label is often a bit dated, it does provide better support from the manufacturer should something go wrong.

    If you’re spraying under hot, dry and windy conditions, the low volume, fine spray approach is irresponsible. Use sufficient water (7 to 12 gpa) to allow low-drift sprays, at least Coarse to Very Coarse, in some case, even coarser.

    Agronomists provide the best possible information for their clients, based on scientific evidence and experience and in accordance with their professional code of ethics. Sometimes the news we deliver aren’t what the customer wants to hear. But we have to represent the interests of all of us, collectively. I find that pretty important.

  • What is Delta T and why is it important for spraying?

    What is Delta T and why is it important for spraying?

    Click here to listen to Audio Article

    Humidity is important in spraying. With the average tank of pesticide being 90 to 99.5% water, evaporation plays an important role in both droplet size and active ingredient concentration. Low humidity causes droplets to evaporate faster, potentially increasing drift and reducing uptake. But relative humidity (RH) isn’t the best way to measure this effect because the same RH at two different temperatures results in two different water evaporation rates.

    Instead, we present Delta T, also known as “wet bulb depression”. Delta T is an atmospheric moisture parameter whose use in spraying has made its way to North America from Australian operations. It is defined as the dry bulb temperature minus the wet bulb temperature, and provides a better indication of water evaporation rate than RH. Higher Delta T means faster water evaporation.

    The recommendations from Australia are to avoid spraying when the Delta T is either too high or too low, with a range of two to eight being described as ideal.

    Figure 1: Delta T chart used in Australia (Source: Australian Gov’t Dept of Meteorology)

    Delta T is being reported on an increasing number of weather stations, and it’s time we took a closer look at what it means.

    Measuring Relative Humidity

    In the early days of weather reporting, relative humidity was calculated from psychrometric charts. All one needed was a hygrometer, usually a sling psychrometer. A sling psychrometer is two identical thermometers side by side whose bulbs could be slung in a circle, exposing them to moving air. One bulb was covered in a cotton wick moistened with distilled water, the other was left exposed and dry.

    Figure 2: Sling psychrometer (Source: ScienceStruck.com)

    As the bulbs met moving air, water evaporated from the cotton wick and that reduced the temperature of that thermometer. The dryer the air, the greater the evaporation rate and therefore the greater the temperature drop. The dry thermometer was unaffected by this movement.

    On measuring the wet and dry bulb temperature, one consulted a psychrometric chart. This chart converted the two temperatures to total water content in the air, compared it to total water-holding capacity, and expressed it as Relative Humidity. Psychrometric charts are useful for many other air parameters such as dew point, vapour pressure, or enthalpy. (Pause briefly to give thanks that we don’t need to know what enthalpy is.)

    Figure 3: Psychrometric Chart (Source: Carrier Corporation)

    Turns out that RH is a poor measure of water evaporation rate. An RH of 24% at 20 C has exactly the same evaporation rate as an RH of 44% at 35 C. That’s why Delta T is the preferred measurement: it’s linearly related to evaporation.

    Note: Modern electronic weather stations don’t need two thermometers to measure air moisture content, and use polymers whose capacitance or resistance changes with atmospheric moisture. Add an internal look-up table, and we have all the information we need.

    Pros and Cons of Water Evaporation

    It’s important to note that our Australian colleagues caution against spraying when water evaporation rate is both too high and too low.

    Too High:

    • Water evaporates rapidly, reducing droplet size and pre-disposing the smaller droplets to drift;
    • Deposited droplets dry quickly, reducing pesticide uptake which is more effective from a wet deposit.

    Too Low:

    • Water doesn’t evaporate, maintaining the smaller droplets in a liquid state. These small droplets are already drift prone, but are now more potent because of more effective uptake. Overnight conditions that are inverted are usually humid, adding to harm potential from the inversion.

    Delta T in North America

    The addition of this parameter to our spraying weather lexicon has been useful. But it’s important to understand the context in which it was developed to properly judge its suitability.

    Aussies started talking about Delta T because the use of finer sprays under the hot dry conditions found during their summer sprays resulted in significant evaporative losses, significantly greater drift potential, and potential reduction of product performance. The guidelines to avoid spraying when Delta T exceeds eight or ten originate there.

    A few changes have happened since these guidelines were developed. Over the past ten to 20 years, we’ve observed greater use of low-drift sprays, with the coarser sprays’ larger droplets resisting fast evaporation. In the past five to ten years, water volumes have increased due to our heavier reliance on fungicides, desiccants, and contact modes of action. Both of these developments have helped reduce the impact of a dry atmosphere. We simply can’t say if a Delta T of 10 is too high with these new application methods.

    Looking at it another way, if Delta T values are very high, increasing water volume and droplet size will mitigate that to some degree, as the Aussies state in their extension materials (linked earlier).

    Formulation

    Pesticide formulation can also play a role in evaporation. Once the water is gone, oily formulations may still have good uptake because the oily active ingredient stays dissolved in the oily solvents. This is both good and bad, helping on-target efficacy but also increasing the risk of more potent drift. Solutions, on the other hand, are more likely to leave their actives stranded on leaves as crystals once the water is gone.

    Bottom Line

    Delta T is definitely useful information when spraying. It will typically rise and fall with air temperature as the day proceeds, and it is wise to consider suspending operations when values are critical. Take note of the Delta T when spraying the same product throughout these hot days and learn from the experience. Remember, the atmosphere affects not just sprays but also plants and insects, and due to this complexity we may not be able to attribute success or failure to just one measurement.

  • Question of the Week: Fine Sprays for Fungicides?

    Question of the Week: Fine Sprays for Fungicides?

    The following question arrived from one of our prairie clients last week:

    “A retailer is promoting the use of hollow cone nozzles to be used on field sprayers (20” spacing) to apply fungicides which he claims out-perform any regular and twin fan tips. Claims:

    • create an extra fine droplet for maximum coverage on the canopy
    • use less water, less time spent filling
    • apply at 3.5 gpa
    • add vegetable oil to reduce drift

    “So his direction to a specific customer was to use the TEEJET CONEJET TXA8001VK nozzle at  80 psi – travelling at 10 – 12 mph to achieve a 3.4 gpa application rate with a ‘very fine’ droplet size.

    “What are your thoughts?”

    Here’s how I answered (edited for clarity):

    That recommendation sounds familiar – it originates from a consultant with experience in South America, where this idea is promoted to improve (aerial) spray productivity.

    I fundamentally disagree with his approach. Adopting and promoting it is not only illegal (contravenes every modern label’s water volume and spray quality requirements), it also puts a generation’s worth of stewardship efforts on drift management at risk.

    To be balanced, let’s explore the attractiveness of this approach. Finer sprays do provide superior coverage and save water. Every child knows this. Finer sprays also go places in the canopy where the coarser sprays can’t, for example very dense lentil canopies.

    Over the years, we’ve explored the performance of fine fungicide sprays in canola, pulses, and cereals in research trials with the U of S and AAFC. To our surprise, droplet size played only a small role in fungicide performance. Water volume was much more important. Droplet size management with pressure through a low-drift nozzle was enough to get the best disease control.

    The main drawbacks of very fine sprays are:

    1. The fine droplets evaporate to dryness very quickly, in seconds. As they shrink, their drift potential is increased even more, and once dry, the remaining particles work much less well. The proponent corrects for this by adding an oily adjuvant as an evaporation retardant. With oil, the fines remain liquid much longer. Although many products become more effective this way, they also become more phytotoxic and less safe for the applicator and bystander. Completely off label, completely risky for crop safety, unknown effects on MRLs, extremely unsafe for the environment and humans. Remember when people dissolved 2,4-D ester in diesel, back in the 40s and 50s and sprayed it with their brass 6501 tips? That’s what this is.
    2. Cone nozzles are designed for airblast sprayers and do not produce good pattern overlaps for boom sprayers. The proponent of this method actually recommends that the boom be raised to overcome the bad patterns and to (believe it or not) simulate aerial application. If this were done, the spray would be re-distributed by air-currents and come down wherever the wind blows it. Probably far away.  The concept of on-target, uniform application, the practice that makes product use acceptable, and the thing we try to achieve with flat fans at a low boom height, is completely lost.
    3. Producers will not have the support of pesticide manufacturers should a performance issue arise. Even worse, if regulators find out about this off-label practice, significant fines (fines for fines, get it?) can be charged under the Pest Control Products Act.
    4. Airborne spray drift with an air-induced spray like the AirMix, GuardianAir, AIXR and the like, applying 10 gpa, is about 1% of the applied amount, measured at 5 m downwind of the downwind edge of the swath in a 20 km/h wind. We’ve never measured hollow cone drift from a boom sprayer, but when we used a flat fan at 5 gpa, drift increased to about 8% of applied. I’d guess a high pressure hollow cone would easily double or triple that. Illegal and irresponsible.
    5. Travel and boom turbulence is a part of faster travel speeds. This would affect the finer droplets much more than the coarser ones, as we can imagine. It’s similar to drift. With a low-drift spray, the proportion of the total spray volume that is “fine”, say less than 150 microns, is about 5%. For a very fine hollow cone, it might by 50 to 75%. So a much greater proportion of the sprayed dosage would be susceptible to uncontrollable movement. This could be good, when turbulence redirects the spray to places that are unreachable by larger droplets. Or it could be bad, as turbulence pushes droplets away from an important target, creating a miss. On balance, bad. Very bad.

    These types of recommendations are concocted by people who want to tell a unique story that is popular with some. Their approach differentiates them from the rest of the crowd, an old and effective marketing trick. But these proponents do not have the best interests of the industry in mind.

    Our individual and collective agricultural practices must be respectful of others. Of safety. Of the law. Of the environment. We have lots of opportunities to make shortcuts…nobody’s watching most of the time. But that doesn’t make it right. It’s certainly not in ag’s long-term interest.

    When considering our agricultural practices, imagine describing them to a young non-farming person. Can you justify your actions? Do your practices make you proud? If not, you have work to do.

    Here’s a task: If your boom sprayer has nozzles that produce very fine sprays, take them off and throw them in the garbage. Might sound radical, but it’s the right thing to do.