Why Do NSF H1-Registered Lubricants Still Allow a 10 ppm Contact Limit Instead of Being Fully Food-Safe?

Last verified: 22 July 2026

"If my lubricant is NSF H1-registered, why does the paperwork still admit that up to 10 ppm of it can legally end up in my food?" That question lands in procurement inboxes more often than most food-plant engineers admit. The short answer: H1 registration was never built to mean food-safe. It is a review of a formulation against a permitted-substance list, and the 10 ppm figure most people quote is not a blanket safety threshold at all — it is a tolerance ceiling written against one specific substance class, mineral oil. The registration and the number answer two different questions, and almost nobody in the supply chain separates them.

Why does a "food-grade" lubricant still carry any contact limit at all?

Because the underlying US rule was never written to prohibit contact. It was written to manage it.

21 CFR 178.3570, titled "Lubricants with incidental food contact," authorises specific lubricants for use on machinery involved in producing, packing, processing, treating, packaging, transporting or holding food — where that lubricant may have incidental contact with the food itself. Not direct contact. Not intended contact. Incidental: a drip point, a splash zone, a chain that runs a few inches above product.

The regulation does not call any of these substances food. It authorises them for a specific mechanical role, under specific limits, on the assumption that some trace contact is physically unavoidable in a running plant.

That is the whole premise of the section, and it explains the discomfort buyers feel when they first read the actual text. A "food-grade" lubricant, under this rule, was never asked to be edible. It was asked to be tolerable in trace amounts.

Where did the 10 ppm figure actually come from — and what does it cover?

This is the part almost every trade article gets wrong, and it is the single fact worth remembering from this whole piece.

10 ppm is not a limit on H1 lubricants generally. It is the specific ceiling the regulation writes against mineral oil: "Addition to food not to exceed 10 parts per million." That is the exact scope of that number — one substance, one ceiling.

Other substances permitted in the same section of the regulation carry entirely different numbers. Dimethylpolysiloxane is capped at "not to exceed 1 part per million" in food — one-tenth of the mineral-oil figure, for a different chemistry entirely. Two more listed substances aren't expressed as food-contact ppm figures at all: aluminum stearoyl benzoyl hydroxide, permitted only as a thickening agent, is limited to 10 percent by weight of the mineral oil, and sodium nitrite, permitted only as a rust preventive, to 3 percent by weight of the mineral oil. Those are formulation ratios, not contamination ceilings.

10 ppm became industry shorthand for "the H1 limit" for one practical reason: most H1-registered products are mineral-oil-based, so most products in the market do carry that number. But treat "10 ppm" as a property of the base oil, not of the registration category. A synthetic H1 lubricant built around a different chemistry may be governed by a different figure in the same table — or by none of the food-contact ppm figures at all, depending on what it's built from.

One more detail the regulation is silent on, and it matters: the limit is measured as an addition to the finished food — not to the lubricant, not to the equipment surface, not as a percentage of the lubricant container. As written, 21 CFR 178.3570 specifies no testing protocol for confirming that number and no stated consequence for exceeding it. That is a real gap in the text, not an oversight in how this article summarises it.

Here is the sentence worth carrying into your next supplier call: the number describes what may be tolerated, not what is safe.

Who assigned the H1 letter, and why does it trace back to a program that no longer exists?

Here's the part most buyers have never been told, and it changes how much weight the letter alone should carry.

The H1/H2/H3 category letters did not originate with NSF. They descend from a US Department of Agriculture program — its Compounds and Packaging Review Branch (CPRB), which authorised nonfood compounds and proprietary substances for use in federally inspected meat and poultry plants.

That program ended in 1998. A Federal Register notice on 13 February 1998 announced the elimination of the prior-approval scheme, and the CPRB itself was formally eliminated on 30 September 1998. Two dates, one sequence — not a contradiction, a wind-down.

NSF International launched its own Nonfood Compounds Registration and Listing Program in 1999, stepping into the gap the USDA left behind. NSF is a private organisation. It inherited the letter system informally, kept it recognisable to the industry, and built a registration process around it — but the letters themselves are the fossil of a discontinued government program, not a live federal classification.

That history is not a criticism of NSF's program. It is context every buyer should have before treating "H1" as a government seal.

What is the actual difference between H1 registration and ISO 21469 certification?

This is the distinction that resolves most of the confusion in this article, and it is worth being blunt about.

NSF H1 registration is a review of ingredients — a formulation and label check against the FDA's permitted-substance rules. The evidence it produces is a listing entry with a registration number attached to a specific product name and manufacturer. It does not audit the plant that made the lubricant. It does not independently test the finished product on a recurring basis. Industry commentary has described it, not unfairly, as running on an honour system — an unaudited self-declaration checked once against a list, rather than an ongoing verification. That is an industry characterisation, not NSF's own language, but it is a fair summary of what the process does and does not include.

ISO 21469:2006, "Safety of machinery — Lubricants with incidental product contact — Hygiene requirements," is a different kind of instrument entirely. It certifies both the product and the process: formulation and label review, a manufacturing risk assessment, an on-site audit of the factory's recordkeeping, quality control and good manufacturing practice, and analytical testing of the finished product. To keep the certification, the site submits to unannounced annual audits and repeat testing. It does not lapse quietly the way a one-time listing can.

Above both sits a layer buyers rarely ask about: who certifies the certifier. ISO/IEC 17065 is the standard a product-certification body must itself be accredited against. ISO/IEC 17025 is the parallel standard for testing and calibration laboratories. Accreditation is what makes a certifier's or a lab's findings worth trusting in the first place — a certificate from an unaccredited body is a document, not proof.

The clean way to hold this in your head: NSF H1 registers a product. ISO 21469 certifies a facility and a process. They are complementary, not interchangeable, and one does not stand in for the other. H1 does not imply ISO 21469. A single product can legitimately carry both. And ISO 21469's own ingredient basis traces back to the same FDA substance-list architecture that underpins H1 — they share a root, even though they check very different things.

Michael Holloway, President of 5th Order Industry, made a version of this argument bluntly in Plant Services on 3 July 2025: he described the 10 ppm figure as a failure threshold rather than a preventive control, and effectively unmeasurable in a working plant — illustrating it as roughly half a teaspoon of lubricant dispersed through a full 55-gallon drum. That is his interpretation, offered in trade press, not a regulatory finding. But his underlying point about the absence of a measurement protocol matches the regulation's own silence on the subject.

Which audits are actually pushing buyers toward food-grade lubricants?

Increasingly, it isn't the FDA rule doing the pushing. It's the customer's own audit scheme.

The BRCGS Global Standard for Food Safety, Issue 8, carried an explicit requirement that materials used for equipment and plant maintenance posing a contact risk — it named lubricating oil specifically — be food grade and of known allergen status. Issue 9 is now the current edition, and its exact wording on this point was not independently confirmed for this article, so confirm the current clause against your own copy of the standard rather than relying on secondhand summaries, including this one.

Other GFSI-recognised schemes reach the same destination by a different route. SQF, IFS and FSSC 22000 address lubricant contact through general maintenance and hazard-control provisions rather than a single named clause; FSSC 22000 connects to it through the ISO 22002-1 series' documented preventive-maintenance requirements. That series was revised in 2025 — ISO 22002-1:2025 now supersedes the older ISO/TS 22002-1:2009 — but FSSC 22000's current Version 6 scheme still runs against the prior text while its own transition to the updated series is underway, so check which edition your buyer's certification body is actually assessing against.

The honest framing is this: food-grade lubricant use is generally a hazard-analysis outcome — a chemical hazard identified at any point where incidental contact is physically possible — which BRCGS turns into an explicit named requirement, while other schemes arrive at the same expectation through their maintenance and hazard-control clauses.

Does the requirement change if you're importing into the EU, Brazil, or Asia-Pacific?

Yes — and the differences are sharper than most import checklists reflect.

In the United States, 21 CFR 178.3570 is the governing text, and NSF's registration program is how the industry operationalises checking a product against it.

In the European Union, there is no EU-wide regulation that names food-grade lubricants specifically. Regulation (EC) No 1935/2004 sets general safety and inertness principles for any material intended to contact food, but it writes no lubricant-specific compositional rule. In practice, EU processors and auditors commonly rely on H1 registration and ISO 21469 certification anyway — by market convention, not by statutory mandate.

Brazil has the clearest confirmed government mandate in this whole comparison. Resolução ANP nº 804, of 20 December 2019, issued by Brazil's Agência Nacional do Petróleo, Gás Natural e Biocombustíveis, governs lubricant registration for commercialisation in the country and — as reported by multiple Brazilian legal and industry sources — requires incidental-food-contact lubricants to be ISO 21469 certified as part of that registration. The ISO-21469-by-name detail was not confirmed word-for-word against the primary ANP text for this article, so verify the current wording directly with ANP before treating it as settled.

In Asia-Pacific, including China, no confirmed government mandate for either H1 or ISO 21469 was found for this article. Where buyers ask for one or both, it is a market expectation, not a legal requirement.

Food-grade status does not change how a shipment is classified at the border, either. These lubricants carry no distinct HS heading — they fall under Chapter 27, heading 2710, alongside other petroleum lubricants. The food-grade status travels in the accompanying documents, not in the tariff code.

What just changed — is there a new number to track?

Yes, and it landed recently enough that many supplier data sheets haven't caught up.

NSF Certification Guideline 537, "PFAS-Free Products for Nonfood Compounds and Food Equipment Materials," was announced 24 March 2025. It confirms a product contains no intentionally added PFAS, no post-consumer recycled material and no intentionally used PFAS additives, with Total Organic Fluorine held below 50 ppm, verified by third-party laboratory testing — and retested annually to keep the certification.

To even qualify, a nonfood compound must first be registered under NSF's Nonfood Compounds program or certified to ISO 21469. Food equipment materials go through a separate path, certified to NSF/ANSI Standard 51.

Notice where this sits structurally: annual retesting, not a one-time review — Guideline 537 follows the certification model of ISO 21469, not the registration model of H1. There's a quiet irony worth naming here too: a second ppm figure, measuring something entirely different, now sits on top of the first one buyers already misread.

Which documents should a buyer actually put in front of a supplier before signing an RFQ?

Not all "food-grade" claims carry equal evidence. It helps to think of them as a ladder, weakest evidence at the bottom.

Ask for rung 3 or rung 4 documentation before an order proceeds, not after. A conversation about how a buyer generally checks a supplier's certification claims, what documents belong in a food-grade purchase order, and how to phrase the request inside an RFQ is worth having early, and it's a natural next step once you've identified which rung your supplier is actually standing on.

Watch for these failure modes — each one is a plausible gap between what a document claims and what it actually proves:

• A registration that has expired but is still quoted on a current data sheet.
• A registration number that belongs to a different product than the one actually shipped.
• A certificate whose scope excludes the specific product or the specific manufacturing site in question.
• "Formulated to meet H1 requirements" language — with no registration number anywhere on the document. This is not the same claim as being registered.
• A certificate issued by a body that cannot be traced back to an accredited certifier under ISO/IEC 17065.

None of these are hypothetical categories invented for this article — they are the kind of gaps worth checking for before a claim is trusted.

Standards referenced above are subject to revision by their issuing bodies. Confirm current wording, edition numbers and clause references directly with FDA, NSF, ISO, BRCGS or ANP, as applicable, before relying on them in a purchase decision or compliance filing.

Frequently asked questions