**4. Discussion**

These results are of importance to stored grain managers because they demonstrate that they essentially have no control in terms of temperature management over the periphery layer of a grain mass. Additionally, temperature values they rely on to make informed aeration control and inventory management decisions heavily depend on the sensor numbers and placement in the rest of the grain mass. Depending on the temperature cable configuration, the average grain temperature reported may be off by several degrees (e.g., 6–12 ◦C in these examples) from the overall temperature average in the grain mass. A key reason for this discrepancy is the fact that grain temperatures in the periphery are lowest during winter and highest during summer but are generally not captured because of the lack of cables placed near the silo wall. The silo wall (and thus a 1 m layer of grain closest to the wall) experiences the greatest temperature fluctuations during a one-year storage period. This result highlights one of the problems of temperature cables which are fixed in place and generally not close to the wall.

This result also demonstrates that it is important to select the sensors used to decide when fans are turned on and off. In this case, only the center sensor was used to make that control decision. The core of the grain mass is not influenced by the weather effects on the periphery, and once cooled during the all-important fall harvest period, the core remains cooler throughout the remainder of the storage period. This could mislead a stored grain manager into thinking that the grain mass has cooled to a sufficiently low temperature during the fall cool-down phase and is as cool as it could get when the rest of the grain mass is still at a higher temperature and not yet sufficiently cooled to mitigate rewarming of grain during the spring and summer storage phase. As a matter of fact, a 4.7 ◦C difference in the average temperature was observed between the low and medium sensor distributions even though the actual conditions were identical across all cases. Activating the aeration fans and operating them to achieve the maximum cooling effect is critical for maintaining stored grain quality and avoiding mold spoilage and insect infestation during the storage period. Thus, relying on the center cable alone is not advisable.

Adding more temperature cables (and sensors) did not provide more useful information than was expected to make informed stored grain management decisions except when temperature readings from the cable in the core of the grain mass was eliminated. The high sensor distribution less the center the cable required six cables and 36 sensors while the medium-plus modified high sensor distribution required nine cables and 54 sensors. For this silo size these additional temperature readings did not provide more useful information for controlling the aeration fan and achieving a lower average temperature than the medium sensor distribution, which utilized only three cables and 18 sensors.

An additional consideration is the number of hours that aeration fans are operated by the automatic controller, which affects cost as a result of electricity consumption. The low sensor distribution had the fewest runtime hours operating for about 20% of the time during the 5-month cool-down and winter holding period (i.e., 741 h out of 3624 h) as shown in Table 2. In comparison, the medium, high, and modified high sensor distributions operated for about 30% of the time, which would be 50% more costly. The modified high plus medium distribution operated the fan for slightly more than a third of the time, and the average temperature of all nodes called for nearly 50% run time which would be 2.5 times costlier than the low sensor distribution. For this U.S. Midwestern Maize Belt location, the recommended fan operating practice consists of three cycles of 150 h during the October through December cool-down period, which results in a fan run time of 20.4% (i.e., 450 h out of 2208 h) [19]. Interestingly, only the low sensor distribution matched the recommended practice relatively over time. However, by allowing fans to operate whenever ambient air was cooler than the average temperature reported by the sensors and the average was above 0 ◦C, fans operated 65–264% more hours (and costlier) than would supposedly be needed based on recommended practice. These findings need to be investigated further in order to refine current aeration decision strategies, especially with regard to the progress of aeration fronts through the grain mass from bottom to top and minimizing fan operating hours and associated electricity costs, which are not considered in this analysis.

In terms of temperatures, these results in Table 3 do not seem to agree closely with those predicted for the three cable-based fixed sensor configurations. For each comparison, the temperature cable results were more than one standard deviation outside the average temperature results for randomized sensor placement. For the low number of sensors case, temperature cable results were more than 4.5 standard deviations (and 2.76 ◦C) above the average temperature for randomized placement. The medium number of sensors case showed the temperature cables were 1.75 standard deviations (and 0.71 ◦C) lower than for randomized placement. For the high number of sensors case, temperature cable results were 1.49 standard deviations (and 0.44 ◦C) above those for randomized placement. The medium distribution case seems to further confirm that excluding sensors placed in the center when deciding whether to turn on or off an aeration fan based on average grain temperature gives the stored grain manager more reliable information to make an aeration control decision than when they are included.

Randomized placement of wireless sensors increases the likelihood that sensors are not placed in the center and that sensors are placed throughout the bulk of the grain mass. This mitigates the core effect by accounting for warmer conditions in the bulk and more variable conditions in the periphery of the grain mass. Randomized placement would result during the filling of a silo assuming wireless sensors are added to the grain stream. Once they hit the grain surface, they will slide or roll at the angle of repose before coming to rest at a location outside the core of the grain mass and sufficient distance away from the silo wall. One disadvantage, however, for not having sensors in the core of the grain mass is that generally stored grain managers do not core the grain mass and remove peaked grain during the harvest season to maximize available storage capacity in a silo. Cooling peaked grain occurs at a much slower rate than the rest of the grain mass due to non-uniform airflow rates. Airflow rates have shown to be almost three times lower through the core of peaked grain than the airflow through the periphery [20].

#### **5. Conclusions**

The MLP 3D finite element model was used to evaluate the effect of number and placement of cable-based and wireless temperature sensors on stored grain aeration decisions and quality management for a 1346 MT (53,000 bushel) silo located in Ames, IA, USA. The key results are:


age, but the predicted average temperatures differed by as much as 2.3 ◦C to 5.9 ◦C compared to the average grain mass temperature predicted by the numerical solution.


**Author Contributions:** Writing—original draft preparation, B.P.; Writing—review and editing, D.M. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

**Disclaimer:** Mention of trade names, proprietary products, or specific equipment does not constitute a guarantee or warranty to the exclusion of other products that may be suitable. It does not imply recommendation or endorsement by the U.S. Department of Agriculture (USDA) or Iowa State University (ISU). USDA and ISU each is an equal opportunity provider and employer.
