banner image on the bottom

Managing stored grain profitably with smart CO2 sensors and AI

calendar icon 16.06.2021
5981

 It has been established with laboratory and field trials that spoiling grain produces higher CO2 levels compared to good quality grain. This is usually attributed to a combination of factors such as the presence of mold or insect infestation. Grain spoilage may occur in localized areas (aka “hotspots”) in a grain store which can be missed by temperature cables installed in a silo, making early detection of spoilage difficult. This often causes costly food waste and adversely impacts the environment as well as agricultural income. Detecting and monitoring CO2 levels, in conjunction with AI methods, offers valuable early warnings about grain spoilage which might otherwise go undetected with conventional monitoring methods [1]. 

It is now possible to use wireless sensor devices to measure small differences of CO2 in the headspace or inside the grain bulk. These differences would be imperceptible to human beings, but comparing them to atmospheric CO2 concentration and combining them with other sensor data, predictive algorithms can detect situations where mold spoilage is developing [2]. This enables growers and grain store managers to take timely corrective action such as aerating, turning, selling, or fumigating the grain. Technology can be used effectively to make grain management easier, more sustainable, and more profitable.

BACKGROUND 

The global average atmospheric carbon dioxide concentration is approximately 410 parts per million (ppm for short). When grain is stored in a silo or a warehouse the main factors that cause the concentration to deviate from atmospheric levels are the respiration of the grain, the fermentation of mold, and the reproduction of insects. In parallel, depending on how airtight the storage structure is, generated CO2 may escape to the ambient environment, causing a reduction of CO2 concentration (Figure 1). 

Figure 1: Schematic of CO2 flows in a silo

CO2 PRODUCTION IN THE GRAIN BULK 

A. Grain Respiration 

Grain is a living and breathing organism. During grain respiration, starch and oxygen are converted to carbon dioxide as well as water and heat. An increase in storage temperature leads to an increase in the respiration rate and thus the CO2 levels (Figure 2). Nutrients being respired lead to losses in the weight and quality of stored produce [3]. 

Figure 2: Representation of the complete combustion (aerobic respiration) of a typical starch carbohydrate

B. Mold respiration 

The growth of toxigenic fungi can adversely affect grain quality and even produce mycotoxins of food safety concern, which should be sensitively monitored and controlled during grain storage. Researchers [4] were able to establish the relationship between the growth of toxigenic fungi and their carbon dioxide production. The results showed the CO2 concentrations increased exponentially during the growth of toxigenic fungi Aspergillus flavus, Penicillium sp. and Aspergillus ochraceus, which was different from the linear increase of CO2 concentration produced by the non-toxigenic xerophilic fungi Aspergillus glaucus and Aspergillus restrictus. The acceleration of CO2 concentration was also found much earlier than the growth of toxigenic fungi, which would be useful for the prevention of grain spoilage (Figure 3). This rapid increase of CO2 concentration in stored grains could be considered as an indication of the growth of toxigenic fungi and consequently of greater risk of microbial spoilage of grains [4].  

Figure 3: Changes in CO2 concentration (blue) and fungal number (red) of toxigenic fungi A. flavus grown at 16% moisture content [4]. The acceleration of CO2 concentration was found to be a precursor of toxigenic fungi. 

C. Insect respiration 

Several scientific studies have used CO2 concentrations as an indicator of insect infestation. More recently, a group of researchers [5] determined the respiration rate of grain insects (S. zeamais, R. dominica, and T. castaneum) at various temperatures. All three species yielded linear increases in CO2 concentrations. The respiration rates of R. dominica were lower than S. zeamais and T. castaneum possibly because R. dominica expresses less movement than the other two species. In the same study, the researchers compared the CO2 production rates in an insect-free silo and an infested silo (filled with paddy rice). As seen in Figure 4, the CO2 concentration curve from the infested silo with higher infestation expressed a steeper slope than the insect-free silo. For those reasons, the researchers concluded that monitoring CO2 concentrations could detect insect presence and potentially be an effective tool for determining insect population density during grain storage. 

Figure 4: CO2 concentration curves in infested and insect-free silos

CO2 SENSOR TECHNOLOGY  

Centaur Internet-of-Crops® provides advanced grain storage technology which includes wireless, battery-operated sensors (Figure 5). These devices measure accurately all the parameters that affect grain quality (such as temperature, relative humidity, O2, and CO2) and fumigation applications (e.g. phosphine). They also transmit safely their data to the cloud in real-time, offering worldwide accessibility. As analyzed above, monitoring gases in a silo offer significant advantages to early spoilage detection since sensors can detect variations in gas concentrations significantly faster than variations in temperature and relative humidity. 

Figure 5: Example of wireless CO2 sensor installation inside a silo

AI-POWERED ANALYTICS OF CO2 IN STORED CROPS

Due to the complexity of mechanisms that produce CO2 in a silo or the way it escapes to the ambient, evaluation of the CO2 readings is not straightforward. Good news is, this now becomes feasible thanks to Machine Learning algorithms. Figure 6 presents the CO2 concentrations of two silo bins as recorded by Centaur sensors. Even though Bin B has higher in-grain CO2 values than Bin A (at least for the first few days), the algorithm does not issue an alert since the grain condition is deemed good. On the contrary, the silo manager of Bin A would receive an alert as early as the 17th day since a sudden rise of the CO2 is detected and flagged by the algorithm as an early sign of grain spoilage.

Figure 6: CO2 concentrations of two silo bins as recorded by wireless sensors. The silo manager of Bin A would receive an alert as early as from the 17th day since a sudden rise of the CO2 is detected. 

 

DATA MANAGEMENT 

Besides monitoring CO2 concentration for grain quality control purposes, it’s also necessary to present data to grain managers or customers in a meaningful and user-friendly way. As part of its Internet-of-Crops® system, Centaur has developed a web platform where all the sensor data is automatically processed and users can receive automated notifications if abnormal behavior is detected. The data in the web platform are updated in real-time, 24/7 and users could access the information from their smart devices at any place. Connecting the system with smart actuators (such as aeration fan relays) is also possible.

Interested in a demo of the Internet-of-Crops® user interface? Contact us.

CONCLUSIONS

CO2 monitoring and analysis can significantly improve grain quality and bring unprecedented benefits for its adopters: 

  • Prevention of mold development and mycotoxin contamination.
  • Prevention of insect development.
  • Reduced fumigation costs from unnecessary fumigation treatments. 
  • Improved grain quality – higher grain market value.
  • Sustainable grain supply chain management.

REFERENCES 

[1] Maier, Dirk & Hulasare, Raj & Qian, B. & Armstrong, P, 2006. Monitoring carbon dioxide levels for early detection of spoilage and pests in stored grain. 9th International Working Conference on Stored Product Protection, 1174, PS10-8-6160 

[2] Francis Fleurat-Lessard, Integrated management of the risks of stored grain spoilage by seedborne fungi and contamination by storage mould mycotoxins – An update, Journal of Stored Products Research, Volume 71, Pages 22-40, 2017. 

[3] Joost Gwinner, R Harnisch, Otto Muck, 1996, Manual on the Prevention of Post-harvest Grain Losses, 2nd Edition, Deutsche Gesellschaft Technische Zusammenarbeit (GTZ) GmbH. 

[4] Huan-Chen Zhai, Shuai-Bing Zhang, Shu-Xia Huang & Jing-Ping Cai (2015) Prevention of toxigenic fungal growth in stored grains by carbon dioxide detection, Food Additives & Contaminants: Part A, 32:4, 596-603 

[5] Chukiat Chotikasatian, Watcharapol Chayaprasert, Siwalak Pathaveerat, A study on the feasibility of quantifying the population density of stored product insects in air-tight grain storage using CO2 concentration measurements, Journal of Stored Products Research, Volume 73, 2017, Pages 21-29. 

scroll-top