BLOG

Digitally Managed Fumigations in the Industry 4.0 Era

03 November 20209 min reading

Vasilis Sotiroudas Pest Management Expert vasilis@centaur.ag

The lack of proper monitoring is not just compromising precision for fumigations and pest control. It actually makes fumigations fail! The problem is that we do not know they are failing, exactly because we are not monitoring properly in order to realize it. This article presents these frequent failure points we have discovered through our work of many years, and goes on to suggest digital and data-driven methods for managing fumigation work towards success.

Large volumes of stored products are fumigated yearly to protect them against pests. Phosphine, sulfuryl fluoride, CO2, low oxygen and other methods are used globally in the fight against stored product insects. The difficulties of accessing silos, warehouses and ship holds, the danger of toxic fumigant gases on human health and the remote location of most storage assets, make fumigation monitoring difficult to achieve. During the COVID-19 pandemic, these challenges have become greater due to restrictions in personnel mobility and access to terminals and storage sites. The lack of proper monitoring is not just compromising precision for fumigations and pest control. It actually makes fumigations fail! The problem is that we do not know they are failing, exactly because we are not monitoring properly in order to realize it. This article presents these frequent failure points we have discovered through our work of many years, and goes on to suggest digital and data-driven methods for managing fumigation work towards success.

As the global economy is embracing the transformative effects of ‘Industry 4.0’ through the use of smart digital devices and data management tools, we hereby recommend a similar paradigm shift for the business sector of commodity and food pest management.

Leakage Almost every storage structure is leaking. Differences in day-night temperatures create internal currents and pressure that increase leakage. A windy day will also dramatically increase fumigant losses. Leakage is expected through the silo roof and silo base, through non-air-tight silo valves, through container doors with aged sealing, through container floors, through ship hold covers with aged sealing, through plastic covers of improper thickness in stack fumigations, through doors and windows, through panel connections or brick walls, through railcar walls, through tarpaulin punctures and floor cracks. See Figure 1 for illustration.

A silo is truly sealed if it passes a five-minute half-life pressure test, for instance according to the Australian Standard AS2628. This is a good method to determine the level of tightness of a structure before applying fumigant. The use of measuring equipment will validate the good sealing. Wireless fumigant sensors that can be placed inside the fumigated volume, such as those available by Centaur, are the best tool for monitoring phosphine, CO2 or low O2 treatments.

Phosphine fluctuation Fumigator perception used to be that product temperature is important only to determine which fumigation protocol can be applied. For example, the Coresta protocol requires 200 ppm for 4 days when temperature is above 20 degC and 300 ppm for 6 days when temperature is between 16-20 degC. A relatively new discovery is that phosphine concentration fluctuates constantly, following temperature fluctuations in the storage area and the laws of gases and thermodynamics. See Figure 2 and Figure 3.

Figure 2: Phopshine concentration dropping when temperature is above 20 degC, in a stack fumigation.

This fluctuation can be as wide as of 500 ppm in 24 hours, dropping concentration from 550 ppm to 50 ppm and turning the fumigation into a certain failure.

Figure 3: Phosphine concentration fluctuating between 850 and 450 ppm during a 12-day treatment, tracking the day-night temperature.

Gas equilibrium Fumigators tend to believe that gas equilibrium is achieved automatically in most fumigations. This can be true only in very small volumes like in containers. Most other storage volumes like silos, warehouses, grain bags, bunkers, sheds, ship holds and large stack fumigations require specific actions to achieve and maintain gas equilibrium. Forced gas recirculation (also called ‘j-system’) or silo-thermo-siphoning are ways to help gas distribution through grain mass. In most fumigations the monitoring points are few and the frequency of measuring is once a day. Under these terms it is impossible to capture fumigant fluctuation and thus impossible to test if the equilibrium is achieved and maintained. Forced fumigant recirculation creates overpressure in the bottom of a structure and under-pressure towards the top of the structure. Both pressure differences lead to leakage. Fumigators realize the leakage (e.g. smell the gas around the fumigated area) and often stop recirculation to limit the leakage. Interruption of the recirculation leads to immediate equilibrium loss and leaves areas of the grain untreated. Figure 4 is a good illustration of all this.

Figure 4: The black vertical line is the moment when gas recirculation is stopped. Equilibrium is immediately lost and product parts remain untreated as phosphine concentration drops below 200 ppm on the third day of the treatment.

Use of wireless sensors has proved that gas concentration varies between higher and lower points in a stack fumigation too. Phosphine is heavier than air and moves near the floor while higher areas get lower readings. The day-night fluctuation is observed in stack fumigations, especially when these happen in a warehouse lacking temperature control.

Sometimes even forced recirculation is not enough to achieve gas equilibrium. Density differences in the product mass direct air currents through the easiest paths, leaving some parts of the mass with less gas (Figure 5).

Figure 5: Even with gas recirculation on from the beginning of the treatment, this fumigation never reaches full equilibrium in a 3000-ton metal silo.

Plan a fumigation - or predict it? Good planning has always been part of successful fumigation. But nowadays Artificial Intelligence and Industry 4.0 models allow the fumigator to predict the treatment and make adjustments in a virtual environment, even before opening a fumigant flask! Fumigation algorithms take into consideration multiple inputs given by the user but they also take data from the meteorological forecast and local weather history. The behavior of every silo or warehouse regarding leakage is recorded and utilized to predict future treatments. The recommended dosage depends on insect species and even resistance statistics in the specific region. Figure 7 is an actual screenshot from the Internet-of-Crops® cloud app by Centaur, which overlays a predictive model with actual measurement data coming from phosphine sensors. The successful termination of a treatment is predicted, while deviations from the model are flagged immediately so that corrective action can be taken.

Figure 6: A stack fumigation with 20 measuring points and a measuring frequency of 2 hours, reveals large differences between areas and specific drops following the day-night temperature fluctuation. Another unexpected part of this graph is around March 28th (day 2 of the fumigation), when in the same stack, at the same moment, there are readings of 4,000 ppm and 200 ppm!

Can a failing fumigation be saved? A corrective action can be simply a timely gas top-up (Figure 8). In several fumigations, adding gas during the treatment is possible, for instance by adding more phosphides in tablet or pellet form. This is even easier when cylinderized gases are used or when a generator is involved. The need for adding gas during a treatment is usually mandatory when the preparation has not been so thorough, as can happen due to time pressure or other logistical constraints. Then leakage can be more serious than expected and fumigant concentration will drop fast. Continuous monitoring and real time alerts are crucial to save the treatment before some areas in the fumigated area drop below the protocol threshold. Wireless fumigant sensors make real time monitoring very easy, with zero time needed to set up and no tubes or hoses involved! (see Figure 9).

Figure 9: An operator places a wireless fumigant sensor in a container, at the start of a fumigation.

People safety Personal protective equipment (PPE), real-time monitoring, visual and sound alarms shall be involved in every fumigation. Life is very valuable to put at risk and modern technology gives us many tools to use. There is no excuse for risking lives.

Figure 7: The Internet-of-Crops® platform monitors fumigations and issues a prediction for each treatment (orange area) ahead of time. When the real readings get uploaded by wireless sensors, they are displayed as curves against the model (blue and red in this case).

The modern approach says that the perimeter is monitored 24/7 by an autonomous system, reading the conditions every second and sending visual and sound alarms (Figure 10). Professionals visiting the area may carry PPE monitors, but the perimeter can nowadays be guarded by an independent system. A system that will also help to identify leakages and restore them on time.

Find the truth in fumigation and aeration People believe that when a fumigant is applied the work is done. And when the door opens, aeration is completed. Sometimes we suspect that something in our procedures may not be so perfect, but we are afraid to dig deeper because we are afraid we may discover things that may upset our logistics. An extra day of treatment may be needed, or some delay to warm the products, perhaps some extra fumigant may need to be added, or we may realize we need to aerate for longer to avoid putting people safety at risk. But we need to dig. And we need to monitor. And we need to review. The next graph contains 2 good reasons why this “digging” is needed:

Figure 8: Gas top-up is made on this graph, just in time to save the 200 ppm protocol. Corrective actions shall be applied in a timely manner, to avoid underexposure and surviving insects!

On the left side of this graph in Figure 11 we see that the phosphine concentration in the ambient air inside a chamber is 1000 ppm higher than the concentration in the center of a pallet at the same moment in time. This is a pallet of half kilo bags, in a plastic packaging with various layers of plastic around 10-bag-batches, and more plastic around the pallet itself. The several layers of plastic allow only a portion of the available fumigant to penetrate.

Figure 10: A network of phosphine low-range sensors guards the perimeter of a treatment allowing safe approach to workers.

On the right side of the graph we see that during aeration, the sensors reading the ambient air reach zero phosphine concentration very quickly but the sensors inside the pallet show that the gas is kept much longer and at levels non-safe to enter.

Figure 11: Multiple points of interest in a properly monitored fumigation process

All the data revealed in this example are valuable for the fumigator, and critical for personnel safety. A data-driven fumigation methodology is the only sane choice, in this era of Industry 4.0.

The technology for precise and safe fumigations is here. Let’s use it and make it all safe and successful for fumigators and consumers alike!

Articles in Cover Story Category
08 May 20142 min reading

Mill Design and Selection of Location, Building and Machinery

If you operate a flour mill, the last thing you may ask is having a halt in the production as the s...

08 June 20222 min reading

Developments in Grain Handling and Storage

01 April 20232 min reading

Flour and milled grain product additives