Have you ever opened a bag of hazelnuts and noticed a “stale/closed” smell, seen a few darkened kernels, or spotted a suspicious dust? It’s not just a quality issue: when moulds and moisture come into play, the risk can also become a compliance one. In this guide on aflatoxins hazelnuts storage EU rules limits preservation, we bring together the requirements of Regulation (EU) 2023/915 and the most practical side: how to store hazelnuts properly—both in a warehouse and at home—to reduce the risk of non-compliance.
What are the EU limits for aflatoxins in hazelnuts, and how do they apply to different uses (direct consumption vs processing)?
The first practical thing to know is that in the EU there isn’t a single “one-size-fits-all” limit: it depends on how those hazelnuts are intended to be used. The operational reference is Regulation (EU) 2023/915, which sets maximum levels for contaminants in food and includes specific entries for tree nuts, including hazelnuts. (This is EU-wide legislation applicable in all EU Member States, including Italy.)
Source: https://eur-lex.europa.eu/eli/reg/2023/915/2024-07-22/eng
For hazelnuts intended for the final consumer or used as an ingredient (i.e., commercially “ready” products), the typical limits indicated are:
- Aflatoxin B1 (AFB1): 5 µg/kg
- Total aflatoxins (B1+B2+G1+G2): 10 µg/kg
If, instead, the hazelnuts are to be subjected to sorting or other physical treatment (in practice: they are not yet “fit” to be placed on the market or used in recipes, but may become so after cleaning/sorting), the typical limits are higher:
- AFB1: 8 µg/kg
- Total aflatoxins: 15 µg/kg
This distinction really changes things, even outside the lab. If you buy or sell hazelnuts “for processing,” you must be able to demonstrate they are genuinely destined for a physical treatment before reaching consumption or use as an ingredient, and that they are managed separately and traceably. If they are already presented as “ready,” the lower limits apply immediately.
There is also a technical note that often sparks debate: when in-shell hazelnuts are analysed, for calculation purposes the contamination is assumed to be entirely on the edible portion (the kernel). In other words: the result must be interpreted as if everything were concentrated in the part that is eaten, affecting calculations, disputes, and comparisons between buyer and seller test results.
Source: https://eur-lex.europa.eu/eli/reg/2023/915/2024-07-22/eng
A practical example explains more than a thousand words: if a sample tests at 7 µg/kg AFB1, it is non-compliant for “consumption/ingredient” (limit 5), but it could fall within the “to be subjected to physical treatment” category (limit 8) if the intended use is correct and handling is rigorous, with segregation and consistent documentation.
And why is all this so “sensitive” in the market? Because mycotoxins—aflatoxins first and foremost—remain among the most critical hazards for tree nuts, including in terms of alerts and border rejections. That’s one reason why investing in storage and controls isn’t a nice-to-have, but insurance against holds and disputes.
Source: https://www.foodsafetynews.com/2025/05/europe-records-rise-in-food-safety-alerts/
What drives aflatoxin risk during storage: humidity, temperature, ventilation, and storage time
The most practical observation is this: aflatoxins don’t “appear out of nowhere”—they are produced by moulds, especially Aspergillus (section Flavi), which can develop both before harvest and, very often, afterwards, when hazelnuts are dried and stored. If conditions favour fungal growth, the risk increases.
Source: https://www.efsa.europa.eu/en/news/aflatoxins-food-efsa-assesses-new-proposed-maximum-levels-almonds-hazelnuts-and
This is where a concept that matters more than moisture percentage alone comes in: water activity (a_w), i.e., how much water is actually “available” to microorganisms. Two lots with similar moisture can behave differently if a_w differs. That’s why, in practice, it makes sense to measure a_w at intake and then during storage, especially if storage is long or the environment is variable.
Recent literature on hazelnuts stresses a point that sounds obvious but isn’t: drying quality is decisive. “Bad drying” situations—for example slow drying in the shade, in heaps, or followed by re-wetting—increase fungal presence and can lead to very high contamination. In an extreme scenario reported in the literature, market samples showed AFB1 above 450 µg/kg, a number that makes clear how huge the problem can become—economically and reputationally.
Source: https://pmc.ncbi.nlm.nih.gov/articles/PMC12846331/
And temperature? Storing “cold” can help, but it’s not a magic wand. If the product is well dried, temperature may matter less than having avoided re-wetting and damp spots. The practical message is: first secure drying and moisture control, then decide how much cold storage you really need.
Source: https://pmc.ncbi.nlm.nih.gov/articles/PMC12846331/
Ventilation and condensation are the other dangerous pair. Cold points, day–night swings, and humid air entering a warehouse can create condensation on walls, roofs, silos, or containers. A local rehydration is enough to create a “hotspot” where moulds find their perfect microclimate, even if the rest of the bulk looks fine. Here, sensors and anti-condensation procedures aren’t bureaucracy: they’re prevention.
Finally, time: think of risk as “time × conditions.” The more months pass, the more likely something will drift, even due to a single abnormal event. FEFO-type rotation (first expiry, first out) and extra attention to origins or seasons with climatic stress help reduce surprises.
How to set up storage “built to avoid non-compliance”: moisture targets, condensation control, and silo/bin management
The practical question is: “What do I need to decide before I even store?” The answer is: a clear internal specification on what you accept and what you don’t, and what you do if a parameter is borderline. Generally, it makes sense to set intake targets for moisture and/or a_w, alarm thresholds, and criteria for rejection or reclassification—distinguishing between shelled and in-shell hazelnuts, because they react differently to air humidity and handling.
Then there’s the often underestimated part: containers. Silos, bins, and storage areas must be managed like “food-contact” surfaces: dust residues, crusts, and old product invite moulds and pests, and they also complicate traceability. Regular cleaning and preventing cross-contamination between different lots are a concrete form of risk control, not just an audit requirement.
On condensation, you need both a “structural” component and a behavioural one. Insulation and make-up air management help, but daily rules matter too: opening hatches only when it makes sense, allowing time for thermal equalisation, avoiding bringing in humid air. A typical example: night loading with colder air followed by daytime warming can cause condensation on walls or on a silo cone; corrective action must be immediate, because that point can become the source.
Ventilating is not always “good”: it only makes sense if the incoming air is compatible in terms of temperature and relative humidity. Documenting cycles and keeping an eye on simple indicators such as ventilation hours, average humidity, and the internal–external temperature difference helps you understand whether you are drying—or unintentionally wetting.
Finally, watch handling moments: loading, unloading, transfers that break kernels and generate dust increase exposed surface area and can worsen the situation. Sieving and dust aspiration, in addition to improving quality, also reduce micro-niches favourable to moulds. In terms of aflatoxins hazelnuts storage EU rules limits preservation, preventing these details is often more effective than “fixing it” afterwards.
Monitoring and sampling plan: when to test, how many tests to run, and how to interpret results against EU limits
The most practical thing to remember is that aflatoxins are not evenly distributed: you can have a few highly contaminated kernels among many “clean” ones. That’s why sampling is crucial, and the EU addresses it specifically in Commission Implementing Regulation (EU) 2023/2782, which provides guidance on sampling for mycotoxins.
Source: https://eur-lex.europa.eu/eli/reg_impl/2023/2782
For “tree nuts” (excluding pistachios and Brazil nuts), a typical indication is that for lots ≥ 15 t you take at least 25 incremental samples to obtain an aggregate sample of 20 kg. The idea is simple: the more heterogeneous the product, the more you need to “go and find it” in many points—otherwise the result risks being unrepresentative.
When should you sample? In practice, four moments make sense: at intake (ideally before or during unloading), mid-storage if storage is long, before shipment if an analytical certificate is needed, and after abnormal events such as condensation, leaks, or plant downtime.
As for testing, operations often work on two levels: a faster, more frequent screening guided by risk, and a confirmatory analysis (for example using chromatographic techniques) when robustness is needed, when the value is close to the limit, or when the customer requires it. In any case, grinding and homogenising the sample are decisive steps: if you do them poorly, even a good sampling plan loses meaning.
And how do you interpret the result against the limits? The sampling and analysis regulation also references measurement uncertainty and decision rules in official controls: essentially, a lot can be rejected if the exceedance of the limit is confirmed taking uncertainty into account (with regulatory criteria that also consider recovery correction and expanded uncertainty).
Source: https://eur-lex.europa.eu/eli/reg_impl/2023/2782
Finally, don’t forget intended use: “to be treated” and “for consumption/ingredient” are not just labels, but categories with different limits and therefore different decisions. Here again, the compass is useful: aflatoxins hazelnuts storage EU rules limits preservation means the analytical result must be read together with the intended use and the real-world handling.
What to do if a lot exceeds the limits: segregation, downgrading, possible rework, and document management
The first move, when a result exceeds or risks exceeding the limit, is almost obvious but decisive: hold and physically segregate. Quarantine, shipment stop, identification of sub-lots, and system blocking prevent the worst outcome—mixing and “contaminating” a larger mass, turning a manageable issue into a disaster.
Then comes the question that changes the ending: is that product truly intended to be “ready,” or can it be correctly managed as to be subjected to sorting or physical treatment? Here the limits in Reg. (EU) 2023/915 for hazelnuts come back: 5/10 for consumption or ingredient, 8/15 for product to be treated. The choice cannot be made retroactively “just because it’s convenient”: it must be consistent with labelling, documents, and segregated handling.
Source: https://eur-lex.europa.eu/eli/reg/2023/915/2024-07-22/eng
On rework, it is
After rework, you need a new robust sampling, because you must demonstrate the intervention worked. And if it still doesn’t comply? You assess downgrading to permitted destinations, or other solutions provided by applicable legislation and commercial agreements, taking practical costs into account: segregated logistics, dedicated storage, possible disposal.
The documentation side is not optional: recording root cause (drying, condensation, storage), corrective and preventive actions, communications, and updates to the self-control plan is what allows you to explain what happened and, above all, what you changed to prevent it from happening again.
Operational checklist for storage and transport: good practices, records, and critical points for audits and official controls
If I had to sum it all up in one practical observation, it would be this: the difference between “we were careful” and “we can prove it” lies in records. In the warehouse, regular checks on silo and container cleanliness, absence of leaks, pest management, roof and wall integrity, and dust control are the foundation—and photos and signatures genuinely help when someone asks for evidence.
For environmental parameters, having a log of temperature and relative humidity, and if possible probes in the bulk, makes it possible to identify conditions favourable to condensation. Ventilation should also be treated as a “traceable” operation: when you did it, for how long, under what external conditions, and with what results.
During handling, rules are there to prevent breakage and reduce exposure to humid outside air. Cleaning conveyors and elevators and preventing cross-contamination between different lots are details that, taken together, make the difference.
In transport, the checklist is simple but non-negotiable: trucks or containers clean, dry, odour-free, and without residues. On long routes, especially by sea, container condensation is a real risk: clear instructions to the carrier and checks at loading help prevent a product that was well stored at origin from arriving “stressed” at destination.
Finally, sampling and analytical certificates: who takes the sample, with which tools, from which points, how chain of custody is ensured, and how a counter-sample is stored. Regulation (EU) 2023/2782 specifically reminds us how much homogenisation matters for mycotoxins, because heterogeneity is the rule, not the exception.
Source: https://eur-lex.europa.eu/eli/reg_impl/2023/2782
And yes, the market context matters too: alerts on tree nuts and mycotoxins remain a reminder that controls and prevention are under the spotlight.
Source: https://www.foodsafetynews.com/2025/05/europe-records-rise-in-food-safety-alerts/