Preservative check

What it is

On the Validation tab of the recipe editor, inside the Formula validation panel, there is a block titled Preservatives and market limits. It compares the preservatives it can see in your formula against the limit AreaCacao has on record for the market you are selling into and the product type you are formulating.

The block says it in its own opening line: "Deterministic calculation on your formula. No AI involved, no credits used." It is arithmetic on data: the same formula, market and product type always return the same result.

For the on-screen walkthrough, see Using the preservative check.

There is no "green" state

This is the first thing to understand, and it is a design decision rather than a gap: the block has no state that means "approved". There is no sign-off, and no colour that means "you are fine". There are five verdicts and none of them absolves.

Three of those five have a variant with its own wording:

• No preservatives identified — the neutral state when nothing was identified. The block warns that "us not identifying one is not the same as there being none: we only see what your ingredient records contain".
• Above the level the FDA documents — the "close to the limit" variant in the United States. That level is not a ceiling: go past it and showing that the use is still safe becomes the manufacturer's job.
• Additive removed from the list — the "not authorised" variant when the substance was once listed and then removed, with its date. Calcium sorbate (E 203) in the European Union, removed on 12 August 2018, is the case in point.

Beyond the five verdicts, the block has three states that are not a verdict and should not be mistaken for one: Checking, Not checked (the check never ran) and No formula. They are described in the how-to page.

"Below the figure we hold" does not clear you to produce

The neutral state always carries a panel titled What this does not say, with this text:

It does not say you can make and sell this without further checks. The calculation counts the preservative you add yourself, not whatever may arrive inside other ingredients, and that can only add. The limit is the one we have documented as of the date shown, not necessarily the one in force in your market today. Labelling, the food category and every other obligation sit outside this calculation.

Underneath, headed What we checked before showing you this comparison, comes the list of everything that had to be closed before the neutral state could be shown — which is also the list of the only things that were checked:

• That every figure in the calculation is a number and none is missing.
• That the documented limit exists and is a positive figure.
• That the limit row is verified against the text of the rule in force.
• That none of your ingredients is left unchecked.
• That the engine and the screen work out the same figure.

If any one of those five cannot be closed, the block withholds the neutral state: it moves to "Cannot be assessed" and says which check failed.

The ceiling belongs to the GROUP, not to each preservative

This is the classic mistake, and it is counterintuitive. In the European Union, for fillings and coatings, the 1,500 mg/kg ceiling does not belong to each additive: it belongs to the whole E 200-219 group — sorbates, benzoates and p-hydroxybenzoates together — and it applies to the total. The footnote to the rule that the block cites says, in as many words, that the additives may be added individually or in combination and that the maximum level applies to the sum.

A worked example on a 1 kg filling:

Each preservative on its own sits at less than half the ceiling. Added together they are over the 1,500 mg/kg. When that happens the block spells it out: "every preservative on its own stays below the ceiling, but the limit belongs to the group and applies to the total".

Two details that change the arithmetic:

• A group can contain a sub-ceiling. In the European Union the p-hydroxybenzoates (E 214-219) carry their own maximum of 300 mg/kg inside the group's 1,500. The block evaluates both rows and the worse one decides the verdict.
• Group membership changes with the market. In Canada, for instance, benzoates and sorbates have independent budgets of 1,000 mg/kg each: there they are not added to one another. That is why the market is never assumed for you.

The Contribution per preservative table shows a row per substance and the Group total, and where the ceiling is a combined one it adds the note "the limit applies to the total for the group, not to each preservative separately".

Limits are counted as free acid, not as the salt you weigh

You weigh out potassium sorbate. The rule counts sorbic acid. They are not the same figure, and skipping the conversion always errs on the reassuring side.

0.1 % potassium sorbate is not 1,000 mg/kg against the ceiling: it is 746. Ignoring the conversion leaves close to 25 % of the dose out of the count.

Three practical consequences:

• In the contributions table the Concentration column is qualified by Expressed as — sorbic acid, benzoic acid, p-hydroxybenzoic acid, SO₂, propionic acid, the preservative as such or free acid — depending on what that row of the rule counts.
• There is not always a conversion. In the United States the documented usage level for sodium benzoate refers to the salt, so the figure is not converted there. Converting it would be a 15 % error in the dangerous direction.
• When you declare an ingredient's content, the block asks whether the percentage on the label refers to the preservative as such or is already "expressed as the acid". That is precisely this difference, which is why it is never assumed.

What the figure is calculated on: the filling or the finished bonbon

The same dosage yields two different figures depending on what it is measured against, and the market's rule decides that, not you.

• In the European Union the limit applies to the FILLING. The filling falls in the food category that admits these preservatives (decorations, coatings and fillings). Chocolate products are a different category, with a closed list of additives in which no preservative from that range appears.
• This is why diluting into the shell buys you nothing. It is not that the figure drops once you spread it over the whole bonbon: the count that governs is still the one on the filling, and presenting the bonbon as a chocolate product moves you into a category where the additive is not authorised at any dose. The move that looks like it lowers the number is the one that makes the verdict worse.
• Other markets do measure on the product as sold (Chile and Brazil, among those covered). There the block asks for the figure it is missing: Share of the bonbon that is filling. Without it, it will not pronounce, because the result would move with whatever it assumed.
• Where the rule does not state the basis (the United States, Canada and Mexico, in the data loaded), the calculation is done on the phase you formulated — the conservative reading, since a phase is never less concentrated than the bonbon containing it — and both figures are shown. If the two readings fall into different bands, the block highlights neither and does not pronounce.

The What the figure is calculated on section shows both readings, On the filling and On the finished bonbon, and marks which one governs in that market. Where it applies, it adds the crossover: "at this dosage, the formula goes over the limit once the filling exceeds X % of the bonbon".

Markets covered, and what each one can say

The Target market dropdown offers six markets. If your account country is one of them it is proposed, flagged as a suggestion and changeable; if it is not, the block asks and assumes nothing.

The Codex Alimentarius is not selectable: it is an international reference framework, not the rule of any one market. And if your account country is not one of the six, the block tells you so — "we have no limits loaded yet" — instead of picking one for you.

One formula, six different answers

Nothing shows better why the block asks before it calculates. This is a single formula — 1,000 g of filling with cream and potassium sorbate — run through the application changing only the market and the product type:

Four things to read in this table:

1. Doubling the dose from 1 g to 2 g crosses half the scale. The same 1,000 g of filling goes from comfortable to half a point from the European ceiling. Any weighing error crosses it.
2. Declaring "finished bonbon" does not lower the number: it changes the category and makes the verdict worse. In the European Union and in Brazil the assembled bonbon is a category where that preservative is not authorised at any dose. It is the counterintuitive result of the previous section, seen on screen.
3. The same gram figure is comfortable in one place and over the line in another. 2 g of sorbate sits at 99.5% of the European ceiling and at 149.3% of the Canadian one. There is no single answer that holds for every market, which is why the block does not pick one for you.
4. The two "Not evaluable" results are not failures. In the United States sorbate is authorised without a figure: there is nothing to compare against. In Chile the limit is measured on the finished product and the filling percentage is missing. In both cases the block names the reason.

What this calculation does not cover

These limits are not small print: they are the difference between what the block asserts and what it cannot assert.

• Preservative arriving inside other ingredients. "The calculation counts the preservatives you add yourself. It does not count what may arrive inside other ingredients, because that figure is not in their records." What is unknown can only add.
• Sulphites from raw materials. If an ingredient declares them, the block flags it separately: the rule counts SO₂ from every source, and the residual content of those ingredients is not in the total above.
• Colours. Today the calculation covers preservatives and preservatives only.
• Labelling and every other obligation. They sit outside. In the United States the block adds two reminders: that the standard of identity for chocolate carries a closed list of optional ingredients with no preservative on it, and that the preservative must be declared by its common name and function at any level of use.
• Anything that is not on the ingredient records. The calculation only sees what your ingredient records contain. It assumes that couverture, cream, butter, nuts and sugar carry no added preservatives; if any of yours does, it has to be recorded on its sheet.
• The age of the figure. The limit is the one documented as of the date shown on screen. A figure consulted more than twelve months ago stops counting as verified against the text in force, and with that it stops being able to sustain the neutral state.
• The 80 % band is an AreaCacao product decision, not a figure from the rule. It exists because the calculation counts what you add and not what arrives inside other ingredients.

Where each limit comes from

The instrument cited and the consultation date sit right next to the verdict, and at the foot of the block Where this limit comes from gives four things: the instrument and article, the food category with its code, the date it was consulted, and one of these three marks:

• Verified against the original text in force — the only one that can sustain the neutral state.
• Verified against an earlier version of the rule; pending check against the one in force — a limit that may have been lowered cannot produce a reassuring result.
• Taken from a reproduction by the competent authority, not from the original text.

The notice attached to all five states

In all five verdicts, right next to the result, this panel appears:

This is not a regulatory validation. It is the comparison between your formula and a figure of ours, with its consultation date in plain sight. Before you go to market, check it against the text in force in your market.

And under Read the full notice, the long version: AreaCacao works out this comparison from the formula you entered and the limits it has documented, as of the consultation date shown; it is not a certification or a regulatory validation, and it does not replace verification by your quality lead or advice from a food law specialist.

Related