Spice sterilisation: specify the result, not the method

Spice sterilisation is the part of a purchase specification most likely to be written in three words and regretted over three months. The usual three words are "steam sterilised", and the trouble with them is not that they are wrong. It is that they answer a question nobody was asking, while leaving the real one open. What follows is what those words commit a supplier to, what they leave entirely free, and the three lines that would have closed it.

This matters more for spices than for almost anything else a food business buys, for a reason that surprises most buyers when they first check it: there is no legal number to fall back on. For a great many foods, a specification sits on top of a statutory limit and tightens it. For a spice, there is nothing underneath.

Sterilised is the wrong word

Start with the vocabulary, because it sets an expectation the process cannot meet. A sterile product contains no viable organisms at all. Nothing done to a commercial lot of cassia, pepper or star anise achieves that, and nothing is trying to. What these treatments do is reduce a microbial population by some number of orders of magnitude, from whatever it happened to be to something lower.

That distinction is not pedantry, because the two ideas behave differently in a contract. Sterility is a state: a lot either has it or does not. A reduction is a relationship between a starting point nobody measured and an end point somebody did. A treatment applied to heavily contaminated material and the same treatment applied to clean material produce different results, and the certificate that comes back says only that the treatment happened.

So the first thing worth knowing is that no certificate of treatment is a statement about the goods. It is a statement about the process the goods went through. Those are different claims, and only one of them is about what is in the sack.

The EU sets no microbiological limit for your spice

The instrument that would carry such a limit is Regulation 2073/2005 on microbiological criteria for foodstuffs, and it is worth reading rather than assuming. Its food safety criteria run as a numbered list of food categories, from entry 1.1 to entry 1.29. The list covers ready-to-eat foods, minced meat and meat preparations, meat products, mechanically separated meat, gelatine and collagen, cheeses and milk powders, ice cream, egg products, cooked crustaceans and molluscs, live bivalve molluscs, sprouted seeds, precut fruit and vegetables, dried infant formulae, fishery products, fish sauce, fresh poultry meat and sprouts.

No entry is a spice. No entry is a dried aromatic herb. The words do not appear in the regulation at all, and they were not quietly added later either: the amendment made after the last consolidation deals with Listeria monocytogenes in ready-to-eat foods and does not reach this.

The consequence is the whole reason to write this piece. For a spice, a purchase specification is not a stricter version of a legal standard. There is no legal standard for it to be stricter than. Whatever number appears in your document is the only number that exists for that trade, and if your document contains no number then no number exists at all.

This is not a gap in the law so much as a different design. General food law still requires food placed on the market to be safe, and a business still has to run a hazard analysis and justify what it does. What is absent is the specific figure, and the specific figure is exactly what a purchase contract is for.

Three treatments, three different paper trails

Once the number is settled, the question of how a supplier reaches it becomes a real question with real consequences, because the three usual answers leave three different sets of documents behind.

Steam is the plainest of the three. It leaves process records: time, temperature, pressure, batch. Nothing about it needs to be declared to anybody downstream, which is why a buyer can specify it and then never think about it again. That convenience is also its trap, since a process record is evidence that something was done, not evidence of what it achieved.

Irradiation is the one buyers most often assume is exotic, and it is the opposite. The EU maintains a positive list of foods that may be treated with ionising radiation, and that list has exactly one entry on it: dried aromatic herbs, spices and vegetable seasonings, at a maximum overall average absorbed dose of 10 kGy. Spices are not an odd case for irradiation in Europe. Across the whole Union they are the only case.

What comes with it is a declaration. Under the directive governing irradiated food, treated food and treated ingredients have to be indicated as such, and for goods not sold direct to a consumer the indication travels on the documents accompanying the shipment rather than sitting only on a retail label. A buyer who leaves the treatment open, and receives an irradiated lot, has acquired a declaration obligation they did not plan for and may have to carry onto their own finished product.

Ethylene oxide is the third answer and, for goods bound for the EU, it is not an answer at all. The point is made in the recitals of the same directive that authorises irradiation, which record that contamination of dried aromatic herbs, spices and vegetable seasonings "can no longer be treated with fumigants such as ethylene oxide because of the toxic potential of their residues", and that ionising radiation is "an effective means of replacing the said substances". The European authorisation of irradiation for spices exists, in the legislature's own account, because fumigation stopped being available.

That is worth holding onto when a supplier anywhere in the world offers a treated product without naming the treatment. The question is not whether the treatment was legal where it happened. It is whether the goods are lawful where they are going.

Naming a method is not naming a result

Now the two halves can be put together, and the failure becomes concrete. Take a specification whose entire microbiological content is the phrase "steam sterilised". A supplier reads it, steam treats the lot, and ships. Every word of the specification has been honoured.

The lot arrives with a total plate count of whatever it happens to be. Perhaps it is excellent. Perhaps the incoming material was poor and the treatment brought it down to something that is still above what your process can carry. There is nothing to argue about, because the document you both signed contains no threshold, and a claim that the goods are unsatisfactory has nothing to attach itself to. The supplier did what was asked.

The mirror image is just as common and costs more. A buyer writes a tight microbial limit and also writes "steam only". A lot of incoming material would meet the limit comfortably after irradiation and not reliably after steam. The supplier is now bound to a method that cannot dependably deliver the result, and the honest ones will decline the business while the others will ship and hope. Neither outcome is what the clause was for.

Both failures come from the same confusion. The result is what protects your process. The method is what protects your label, your customer declarations and your own compliance position. They are separate concerns, they are answerable by separate evidence, and putting them in one phrase means neither is actually specified.

What to write instead

Three lines do what the one line does not, and none of them is long.

The first is the result. Name the organism or the count, the limit, and the method of analysis that will be used to decide it. A limit without a stated analytical method is an argument waiting to happen, because two laboratories using two methods on the same lot can return different answers and both be right.

The second is the permitted treatments. Not the treatment: the list. Say which methods are acceptable and which are excluded, and say it because of what each one obliges you to declare rather than out of a general preference. If irradiated material would create a declaration you cannot carry, exclude it in the contract and not in a phone call. If steam is the only thing your customer will accept, then the limit has to be one steam can reach.

The third is the evidence. Say what must arrive with each lot: the certificate of analysis against the named limit and method, and a statement of which treatment was applied. This is the line that turns the first two from intentions into something checkable on receipt, and it is the one most often left out.

Written that way, a disagreement becomes a factual question about a lot rather than an argument about what was meant. That is the entire benefit. Our catalog of whole and milled botanicals lists the forms we supply, and where a buyer specifies steam treatment we supply against it, in house or through a treatment partner, on the same three lines described here. What the trade desk will not do is quote against the word "sterilised" alone, because that word does not describe a deliverable.

What a specification cannot fix

It would be a poor article that ended by suggesting three lines solve the problem, so here is what they do not touch.

A microbial limit is verified on a sample, and a sample is not the lot. Contamination in dried botanicals is not evenly distributed, and a sampling plan that ignores that returns a number with more confidence attached to it than it has earned. The specification says what the limit is; it does not make the sample representative.

A treatment also does nothing about the reason the count was high. Microbial load in a spice arrives with harvesting, drying and storage practice, and a lot that needed a heavy reduction to pass is telling you something about its origin that the passing certificate then conceals. Treatment is a remedy, and a remedy applied routinely is a description of an upstream problem.

And none of it speaks to the other things that decide whether a lot is usable: moisture, foreign matter, pesticide residues, species identity. A specification tight on microbiology and silent on the rest is not a tight specification. It is a specification tight in one place.

What the three lines buy is narrower than safety and more useful than comfort. They make a disagreement resolvable. In a trade where the goods arrive weeks after the agreement, and where the person who wrote the specification is rarely the person receiving the pallet, being able to settle a question by reading the contract and the certificate side by side is worth more than any adjective.