Quality of stored grain: building up a decision support system for management and control of quality of malting barley

A survey of quality requirements for malting barley was achieved in order to collect information on current commercial practices in selected European countries (England, Belgium, France, Spain, Denmark, and Portugal). This information was related to the allocation of resources to carry out storage actions proposed by the planification function of the DSS in order to take into account the specificity of quality requirements from different countries. The economic benefits of storing malting barley arise from the change in price of grain during the storage period. This potential gain has to be offset against the risk of losing quality and the cost of the management actions required to maintain the quality. The activities associated with maintaining quality such as cleaning, drying and pest control each have their own cost and resulting benefit. To evaluate the benefit of any such action the farmer requires reliable information on the likely cos of each activity. This result consists in the description of the costs and benefits of these different actions. These specific costs may be taken into account in the planning of the storage strategies delivered by the Decision Support System of global management of Quality of stored malting barley. They include the specific requirements of the different EU countries major producing and processing malting barley.

Development of a classification system (software) for grain insect noise identification on a stand-alone Digital Signal Processor (DSP) board for stored-grain-insects acoustical signature capture and processing. The functions of this software are: - The detection of stored-grain-insect noise at very low population density levels in malting barley grain bulks, - The insect acoustical signature processing by a procedure of identification of the species and of the developmental stage.

Comparative data of the effects of cooling of malting barley after the harvest on commercial quality criteria and especially the germination capacity and the spoilage by storage fungi and related mycotoxins. In using these data, the storage technical routes including a cooling aeration process were designed and introduced into the knowledge base of the DSS for the management of sanitary and safety quality of stored malting barley.

A model for the reciprocal conversion of malting barley Aw into moisture content (mc) was produced that enables to use observed mc recorded by NIR analysers in mathematical models. A relative humidity sensor was tested in five storage bins equipped with temperature probes and filled with malting barley. Three bins were aerated with ambient air and two were not aerated. The moisture content (mc) of grain was measured and also calculated from automatically measured relative humidity and temperature, using the modified Chung-Pfost equation and a new one, the Jacobsen's equation. The Jacobsen equation gives better fit than the modified Chug-Pfost equation. The greatest discrepancy was under adsorption regime (from cooling aeration) conditions.

Conception of the architecture of the different components of a Decision Support System (DSS): - Knowledge Base (KB), - Inference Engine (IE) and Graphical User/Experts Interfaces (GUI), and of the reasoning system enabling a full communication between them. Preserving grain post-harvest necessarily deals with preventing degradation of quality. The strategy of Preventative SGM (Stored Grain Management) focuses on the maintenance of the initial quality of grain during the overall storage process. It is based on four sequential steps: - Assessment of grain quality and condition before storage (the initial quality); - Planning the optimal storage technical routes, taking into account available equipment and permitted actions; - Monitoring of the grain condition during storage and - Reassessment of the storage technical route if the grain condition drifts out of safe storage conditions. To assess the quality of stored barley, human experts based their reasoning on: - Measurements and observations of grain, - Scientific knowledge on the malting barley kernel and on its pests causing spoilage or damage (the biotic factors of deterioration: early germination, insects, moulds and mites), - Experimental knowledge and know-how. To represent all these aspects of knowledge, a specific qualitative modelling incorporating reasoning and integrating quantitative models (existing and those developed in the QualiGrain Project) has been developed. The planning problem is to control the abiotic factors (time, temperature, moisture content, impurities, and insecticide residues) with a view to keeping them within safe storage conditions. The planning method is hierarchical, with two abstraction levels: - Generation of the genric storage plans, - Refining the generic storage plans taking into account the time, the available equipment and permitted actions. An optimal storage technical route provides a plan of the storage actions to execute during stotrage with an estimated safe storage period after each action of the plan. A prototype of the expert system QualiS (copyright) has been developed in Windows (MS) for use on PC. The advice afforded by the decision support system has been validated firstly by human experts and subsequently in pilot scale experiments in Denmark and UK.

A model predicting the duration of safe storage period in relation to grain temperature and Aw was produced. The domain of validity is covering grain condition in Northern and middle Europe. A mathematical model based on the two controlled variables in stored grain bulks: temperature and activity of water, was developed from experimental data on grain (malting barley) respiration rate (O2 consumption and CO2 production. This model enables the calculation of the rise in grain temperature to the limit of 35°C, which corresponds to the threshold for the deterioration of seed viability. It was used to create another model, usable graphically or by computer, and enabling the assessment of the "safe storage life" of malting barley as dependent on moisture content and temperature. The acceptable threshold for viability of malting barley was chosen at 95% but it can be fixed at any specific level in the model.

The correlation between one-day rapid test (chitted grain at 24 h) and Germination Index (3-day test) was stated in using an image analyser. Production of high quality malt requires that the barley have a fast, completely and uniformly germination. Therefore, it is of great importance that good methods for the rapid assessment of germination are available. All the current germination tests are manual and KVL has been working with the development of new tests with which the level and the uniformity of germinated seeds can be automatically determined. Measuring biophotons with a photomultiplier has been experimented on raw barley samples, but no correlation to reference values was observed. So, KVL has carried out a work with an image analyser to determine germination level and characteristics automatically. The results obtained on a limited number of malting barley samples are promising and a pre-development study was undertaken in order to produce sufficient data to feed the database on malting barley germination required for the adaptation of the existing prototype for grain image analysis to malting barley germination determination.

Software programmes for the four parts of the DSS dedicated to stored malting barley quality management: - The Grain Quality assessment module; - The storage scenarios planification module; - The explanation module about the advised storage routes, and - The GUI condition monitor module. "Quality of stored grain (QualiGrain): building up a decision support system for management and control of quality of malting barley" A prototype of a decision support system based on expert knowledge input, running on Windows on a PC, was developed. This computerised system is a new tool, which should facilitate the transfer of technical and scientific knowledge about stored malting barley quality management to the storekeepers. The prototype DSS supports the user for the choice of optimal storage itineraries (the safest storage route). The reasoning process is based on quality change prediction during storage. It includes four phases: - Assessment of initial quality level of the grain; - Planning the safest storage route; - Permanent monitoring of grain condition in using specifically adapted equipments (acoustic probes, temperature/relativehumidity probes, remote aeration control); - Re-planning the storage route when a deviation from the predicted condition occurs. The reliability of the advice afforded by the DSS has been validated by storage specialists before and after pilot-scale experiments carried out in different European countries. No computer-assisted tool for whole quality management of stored grain did exist before QualiGrain EU Project started. Two important and original results were obtained in this domain during the project: - A computer assisted DSS for quality - assessment, -prevision of changes, and -monitoring; - A knowledge representation on the changes of all quality characteristics together of malting barley (the whole grain bulk excosystem modelling). In its actual shape, the DSS software protoype can be easily used by any grain storage manager having a PC at his disposal for training and demonstration. This copyrighted software is available for further development (e.g. by a software releaser). Among the modifications to be introduced in a commercial software, some improvements for a more friendliness presentation of the outputs on the GUI interface are necessary (most of them are simplifications and shortening). The market sector interested in the use of this PC-assisted DSS is the industry of grain and pulse seeds storage and handling sector. The first-processing industries of cereals and pulses should also be potential users. After the modifications required transforming it into a commercial product, this software should be of interest for software releasers of the agricultural and agro-food industry sector. Technical extension services in cereal production and storage industries may use it for training grain store managers or for the enforcement of good storage practices following HACCP guidelines.

Mathematical models of dormancy break based on initial condition of harvested grain (temperature, activity of water (Aw) and initial germination rate). A specific algorithm was developed for the prediction of the time required to reach the maximum germination rate and the moment when germination rate is reaching the 95% lower limit. Experimental trials on samples of barley collected from 1998 and 1999 harvests were used to investigate three important features of malting barley seed "life": - Dormancy lasting period; - Viability rate change with time; - The storage time period allowed before reaching the commercial threshold for viability (95%). The dormancy experiments enabled the existing model of dormancy to be extended to five further varieties. Work on viability enabled the identification of a typical worst-case scenario model of viability. The dormancy and viability models are combined in the DSS, providing a model that can predict the complete germination history (germinative energy versus storage time) of stored malting barley.

Several models of the fate and decrease in efficacy of several organophosphate (OP) compounds (post-harvest registered insecticides), were produced (especially dichlorvos for which data were lacking at the beginning of the Project). There are several issues relating to pesticide application on stored grain. Thus, it is of prime importance that residues do not exceed permitted levels. So, it is necessary to know decay rates so it can be predicted at any time during the storage period, and in dependence to stored grain condition, what is the concentration and the residual efficacy levels of these pesticides. This result collates work to define residues half-lives and protection periods for the most-widely used pesticide, pyrimiphos-methyl (P-M) and reports experiments on residue and efficacy decline of P-M and dichlorvos at combinations of high and low temperature and high and low moisture, which are the most extreme likely to be experienced under European storage conditions. Experiments on the decline of P-M residues during pilot-scale malting are also reported.

The absence of a significant impact of fumigation by phosphine gas, especially in extreme conditions of stored-grain temperature (10°C and 35°C), on the quality criteria of malt and beer during the whole processes of malting and brewing, was stated. The lethal exposure for complete control of various insect and mites pests of malting barley either with controlled atmosphere or phosphine fumigant were assessed. The experimental trials were carried out for the temperature range between 10 and 30°C on different varieties of malting barley from the 1998 and 1999 harvests. Even with high carbon dioxide exposures, controlled atmosphere disinfestation did not affect neither the quality criteria nor the viability level of malting barley. With phosphine fumigation during one week at 2 g/m3 PH3, neither germination nor the malting did not reveal any significant decay in quality compared to unfumigated samples.

Edited Proceedings of a scientific and technical meeting held in Reims on May 28-29, 2001: This meeting aimed at the communicating the most important results obtained during the Project's work and the new approach of the management of all expected qualities of stored malting barley by a computer-aided DSS. Summary of Proceedings' content: "Quality of stored grain (QualiGrain): building up a decision support system for management and control of quality of malting barley" A prototype of a decision support system based on expert knowledge input, running on Windows on a PC, was developed. This computerised system is a new tool, which should facilitate the transfer of technical and scientific knowledge about stored malting barley quality management to the storekeepers. The prototype DSS supports the user for the choice of optimal storage itineraries (the safest storage route). The reasoning process is based on quality change prediction during storage. It includes four phases: - Assessment of initial quality level of the grain; - Planning the safest storage route; - Permanent monitoring of grain condition in using specifically adapted equipments (acoustic probes, temperature/relative humidity probes, remote aeration control); - Re-planning the storage route when a deviation from the predicted condition occurs. The reliability of the advice afforded by the DSS has been validated by storage specialists before and after pilot-scale experiments carried out in different European countries.

Prototype equipment for insect noise detection inside a grain bulk equipped with three acoustic probes that were designed in order to be introduced in the grain bulk from the surface up to a depth of 2.5m. The probes are individually connectable to a data logger, itself connected to a classification system. This classification system is designed for automatic identification of either the insect stage (larva or adult) and the insect species or family (weevil, grain borer, and external feeder beetle). The classifier includes a reference database of typical acoustical signature of insect species and stages for the major grain pest species. The classification software is running on Windows on a PC. This new system of insect detection at very low level of population density based on acoustic signal capture and processing is an entirely new tool enabling insect infestation monitoring in stored grain bulks. The development and the validation of feasibility for use of this technology is important for economic reasons: It will be possible to detect the presence of insect in grain bulks at a very low level of population density, i.e. a long time before this infestation will become a detectable defect at the inspection for commercial purpose. This new tool is allowing an easier management and in most cases the prevention of an insect infestation before it becomes a storage issue of economic importance. This will promote a change in actual practices of insect control using pesticide applications generating long-time-lasting residues.

A highly significant correlation was stated between ELISA rapid testing analysis and standard analysis (HPLC) for the quantification of the storage mycotoxin: Ochratoxin A (OTA). Taking into account the safety and the quality of the end products of malting barley processing, there is a great interest in a rapid method for the detection of storage mycotoxins and gushing tendency in malting barley. Rapid methods such as the ochratoxin A (OTA) Elisa test and gushing test on malting barley are promising but require further enhancement to be accepted for routine use by grain storage managers. Actual rapid methods enabling to detect fungal spoilage or quality damage from storage fungi in stored barley stocks, do not allow to screen a large number of samples in a short time, as required by the grain handling high flow rates at the loading or the unloading of grain stores. This result is highlighting the correlation of OTA test with the results of a reference method (HPLC analysis). The correlation coefficient is highly significant for OTA extracted either from grain or from malt.

Model of the rate of increase of insect populations on malting barley related to grain Aw and temperature, with a link to the impact of high levels of infestation on changes in malt and beer quality criteria. In using published results and previous work on population dynamics of insects in wheat, predictive models of population dynamics of harmful insects for barley were established. The application of these models to current malting barley storage was examined using laboratory and pilot scale experimental studies on the two main insect species that represent a risk of quality deterioration in stored malting barley: the granary weevil, Sitophilus granrius, and the saw-toothed grain beetle, Oryzaephilus surinamensis. The models were established within a range of temperature and grain moisture content of 15-30°C and 12.5-15.3% (wet basis), respectively. The pilot scale trial of malting barley storage lasted 8 months with the two conditions: infested batches compared to the uninfested ones. Quality criteria of grain and malt were assessed immediately before bin loading and at the end of the storage period. Excepted in one situation with high moisture content (15.3%) and infested grain batch, both germination and all quality criteria of malt were not affected during the storage period. These results were used in the DSS knowledge base in order to take into account the risk of damage of stored product insects on the quality of seeds or malt in the planning of optimal storage route.

Reliable statistical models for the prediction of the impact of heat treatments on germination capacity and dormancy break of malting barley (validated). Mathematical models were developed for the drying performance and for the loss of viability of freshly harvested malting barley, during a quick heat shock or a complete drying process. Reliable statistical models for the prediction of the impact of these heat treatments on germination capacity and dormancy break of malting barley were validated and published in the proceedings of the last meeting of the Project.

The fate of organophosphates (registered for seeds treatment in several European countries), during the malting and brewing processes was accurately stated. Pirimiphos-methyl was applied to two identical batches of barley at a rate of 10mg/kg using pilot malting drums. After one week, each batch was malted, one to give an ale malt and the other to produce a lager. Samples were taken throughout processing and analysed for residues of pirimiphos-methyl. About half of the pirimiphos-methyl applied was taken up by the barley. Most of the remaining chemical was washed out of the drums with the first steep water. Approximately 8% of the residues were lost during steeping. Up to 24% could be lost during germination but the greatest losses occurred during kilning (about 43%). The finished malt contained less than 20% of the residues of the starting barley. The amount of pesticide residue in the beer is such that, only if all survived the fermenting process, would it come within normal residue detection parameters.

Deliverables

Share this page

Download