IQF Broccoli for the United States: Product Controls and a Buyer Supply Plan

The U.S. case begins at the field, not at the export carton. The purchaser's legal and FSVP teams need a source they can evaluate; production needs measurable crown, stem, colour, texture and foreign-material targets; the laboratory needs a residue definition and sample basis; and the broker needs a true commodity description. We propose a release decision that joins these records without confusing a negotiated appearance grade with a federal import tolerance. Blanching and freezing preserve a useful food, but they do not themselves validate pathogen destruction or make a product ready to eat.

Fix the broccoli form, cooking use and contract

We ask for a signed item sheet before offering a lot. It should state the botanical identity, China origin, single ingredient, floret versus spear or cut form, whether crowns and stems are both permitted, dimensions measured after freezing, maximum stem proportion, minimum crown proportion, degree of blanching, expected cooking application, packaging weight, frozen storage and lot coding. The buyer should say whether the broccoli enters foodservice kitchens, a cooked ready-meal line or retail repacking. Those uses change the acceptable visual range, instruction wording and the customer's own food-safety responsibilities.

"IQF broccoli florets" is too short to accept against. One purchaser may need uniform small florets for a vegetable blend; another may value larger crowns for plated sides. A usable drawing or reference sample defines where the cut stem ends and where branching starts. A caliper or sieve result should say whether measurement is by longest dimension, crown diameter or weight class. We would agree the sample size and number of cartons across the lot, count the defects in that sample, and retain a sealed reference or photo set. A good-looking handful from the top of one carton cannot represent a container. The actual GreenLand frozen broccoli product page shows the form, but the buyer's order specification decides the tolerances.

GreenLand frozen broccoli florets showing their crown and cut stem form.

The product is labelled and sold for thorough cooking before consumption. We would place the handling and cooking statement where the industrial customer can carry it into its manufacturing instructions. FDA's ready-to-cook advice says partially cooked food may still contain harmful bacteria and should not merely be warmed. If the customer proposes to thaw it into an uncooked salad or garnish, we pause the transaction to reassess process validation, microbiological controls, intended-use label and supplier agreement. Calling the blanch step "fully cooked" to gain a broader market would create a safety mismatch.

Packing is also a functional decision. A lined bulk carton should protect against moisture pickup, odor and handling damage, with an agreed net weight that fits receiving and production. We should check liner seal, carton closure, pallet pattern and whether a mixed load changes stacking or door temperature exposure. The contract can specify a commercial frozen target such as −18°C, whether the buyer requires a data logger, and what temperature is recorded at loading and arrival. Separately, the current APHIS ACIR frozen-Brassicaceae record requires quick freezing before shipment and commodity temperature never above 20°F (−6.66°C); if above that threshold on arrival, APHIS regulates it as fresh, unfrozen produce. The 7 CFR 319.56-2 definition additionally requires at least 48 hours at −6.7°C (20°F) or below before release. IQF, frozen-storage and transit records should together substantiate that continuous period; it need not all be spent at the U.S. port. That regulatory threshold is not a suggested commercial storage target. A frozen label is a handling requirement; it is not proof that every pallet stayed at the set point. The physical and documentary evidence must carry the same lot identifiers through loading.

Apply current U.S. pesticide tolerances to the source plan

Pesticide numbers must be drawn from the current U.S. commodity and chemical entry, including the regulation's residue definition. The current 40 CFR Part 180 is the source; an old spreadsheet can miss a 2026 revision. As a worked example, the 30 June 2026 EPA rule changed bifenthrin for "Vegetable, brassica, head and stem, group 5–16, except cabbage" to 0.9 ppm, equivalent to 0.9 mg/kg, and removed the old 5A entry. Broccoli is in that head-and-stem crop group. The compliance residue under 40 CFR 180.442 is measured as bifenthrin itself while the tolerance text addresses its metabolites and degradates. A lab's broad "pyrethroids" screen is not the same analytical definition.

Another current example shows why the analyst must read each substance independently. 40 CFR 180.438 lists the Brassica head-and-stem subgroup 5A at 0.4 ppm for the applicable lambda-cyhalothrin entry, with compliance determined from the specified combined lambda-cyhalothrin and epimer residues. One chemical uses a revised 5–16 group while another retains 5A. We cannot substitute a generic "broccoli MRL" or use an EU number. Before placing a raw-material purchase, we and the customer's FSVP team should build an analyte-by-analyte matrix for the pesticides actually used or reasonably foreseeable, with each current section, crop-group membership, residue definition, unit, laboratory reporting limit and test frequency. The examples below are reference checks, not a complete panel or a promise that these chemicals were used in the field.

Illustrative scene: Broccoli florets placed in clean bags for representative pesticide-residue sampling.

Freezing needs a separate legal interpretation. Under 40 CFR 180.1(d)–(e), a frozen food is processed for this purpose. Where no separate processed-food tolerance applies, the regulatory carryover conditions include compliant raw agricultural commodity, good manufacturing practice that removes residue to the extent possible, and no greater residue concentration in the processed food than the raw-commodity tolerance. We therefore examine both the harvested broccoli and the frozen finished lot; we do not assert that blanching removes all residues or that an over-tolerance raw lot becomes lawful because it passed through the freezer. A processing factor, if used, requires commodity- and analyte-specific evidence and an explicit calculation. An analyte with no applicable U.S. tolerance or exemption cannot simply be given the EU default of 0.01 mg/kg.

Heavy metals deserve the same honesty. We have not identified a universal U.S. broccoli-specific maximum for lead, cadmium, arsenic or mercury that can be copied into every purchase contract. The customer's risk assessment may call for a test panel and a negotiated action limit based on source and intended consumers, with appropriate FDA and state requirements checked for that product. We will not import an EU maximum or print a fabricated "FDA heavy-metal limit" in a COA. Sampling method, dry versus wet basis, analytical method and reporting limit belong beside any number the buyer actually agrees.

Control raw broccoli, blanching, cooling and IQF

Raw-material acceptance should join farm identity, crop and harvest dates, pesticide application records, water and field-risk review, transport condition and a visual intake assessment. The grower and exporter can have different legal names; both need traceability. We would define a lot boundary that survives cutting, blanching, freezing and packing, and record any commingling. If the buyer asks for organic, the certified operator, scope, chain of custody and U.S. import certificate route must be reviewed separately. Conventional material does not become organic because no pesticide was detected in one test.

At trimming and cutting, the aim is consistent size without hiding fibrous stems or insect damage. We set a sampling method for crown/stem ratio and pieces outside the agreed size band. Wash water and contact surfaces are managed through documented sanitation and monitoring. A pre-freeze visual sort can remove yellowing, brown damage and stray leaves. Optical colour or shape sorting is useful for visible defects when installed and validated for the product; we do not claim that an X-ray machine sees green-versus-yellow colour. X-ray responds to density and product geometry and may help with specific dense foreign materials under validated settings, while metal detection responds to metal. Neither replaces an inspection of vegetable tissue, soft debris or all insects. The equipment listed on a quote must reflect the actual source line rather than an illustrative photograph.

Illustrative broccoli blanching and cooling sequence, not a photograph of the source facility.

Blanching is set around enzyme inactivation and the customer's texture target. Too little treatment can leave enzyme activity and colour deterioration; too much can make the floret soft and increase fragments. We propose recording time, product and medium temperature, loading depth and any peroxidase or other agreed verification method, then cooling promptly to reduce further cooking. Those numbers are process-specific, so this public article does not invent a universal "90 seconds at 95°C" standard. A plant must validate the operating window for its cut size and line, and retain batch evidence. Blanching is not a demonstrated pathogen kill step just because the word "hot" appears in the flow chart. A lethality claim would require appropriate scientific validation of the exact time-temperature distribution and worst-case product location.

IQF should yield discrete pieces after freezing, with agreed allowance for clusters and broken crowns. Packaging occurs under sanitary conditions after cooling and freezing. The post-blanch zone can reintroduce organisms through equipment, water, personnel or the environment; a blanch log by itself does not close that risk. We ask for sanitation verification, maintenance records, foreign-material checks and temperature records tied to the lot. Frozen storage limits growth, but FDA's food-safety advice makes clear that freezing does not reliably kill pathogens. A thawed sample for quality evaluation should not be repacked as saleable food.

We organize the proposed production flow into six recorded decisions:

1. Approve the field and pesticide record, then receive and inspect the harvest lot.
2. Trim, cut and wash to the agreed size and sanitation plan.
3. Blanch to a validated quality window and any separately validated safety function, then cool promptly.
4. IQF, sort visible defects and foreign matter, and pack hygienically.
5. Hold the lot while production, sanitation, temperature and laboratory evidence is verified.
6. Release the accepted lot and dispatch it frozen with matching documents.

At each transition, a responsible operator needs a record and an escalation path for deviation. If a temperature monitor fails, the decision is based on the remaining credible evidence and the food-safety plan, not an automatic assumption that the lot is safe because it still looks green.

Our proposed detector check records the material and size of each validated test piece, product and pack setting, time, operator and confirmed rejection. Checks cover start-up, pack or product changes, relevant stoppages and the scheduled interval established for that line. A failed check stops release of the affected output from the last satisfactory check; the quality lead defines the lot boundary, verifies recovery and documents disposition before packing or dispatch resumes. Visible colour, size and tissue defects follow a separate grading record.

Separate NRTE controls from RTE claims

For this case, the intended use is not ready to eat (NRTE): the buyer cooks the broccoli. That is a product and channel assumption to be written on the specification and label. FDA's 21 CFR Part 117 preventive-controls framework still requires the covered facility to analyze known or reasonably foreseeable biological, chemical and physical hazards and identify those requiring preventive controls as applicable. The controls, monitoring, corrective actions and verification follow that analysis. NRTE status does not erase hygiene, Salmonella or Listeria concerns, nor does it make a lot with a known serious hazard acceptable.

The precise regulatory distinction matters. FDA says environmental monitoring is required where contamination of an RTE food with an environmental pathogen is a hazard requiring a preventive control. That statement cannot be mechanically imposed as the same legal trigger on every cooking-required frozen vegetable. A facility may still use environmental monitoring as a useful verification activity, especially around wet post-blanch areas, based on its hazard analysis and customer agreement. If the buyer's process no longer cooks the ingredient, the RTE analysis changes and the post-process environment becomes central. The customer's FSVP review should understand that use instead of accepting a generic HACCP certificate as a complete risk assessment.

Illustrative scene: Broccoli on a processing line with visual and foreign-matter checks.

There is no single U.S. statutory total plate count number for all frozen broccoli that we can responsibly quote. A customer may set aerobic count, coliform, Enterobacteriaceae, yeast/mould or pathogen specifications, but method, sample size, number of sample units, acceptance number and action after a positive must be negotiated with its intended use. A certificate of analysis needs the organism or indicator, method, result, units, detection limit and lot. "Passed microbiology" without those fields is not an auditable result. When a pathogen is found or a food-safety control fails, the lot is held while the qualified team investigates scope and disposition; a negative retest of one small sample does not automatically release the rest.

Cooking directions are part of risk communication, yet a supplier cannot outsource every control to the buyer's stove. We would specify fully cooked use, preserve storage and lot coding, and give the buyer the process and test records it needs for FSVP. We would avoid a "washed, ready to eat" icon or a picture of raw broccoli salad on this SKU. If the customer's prepared meal has a final validated cooking step, its own food-safety plan should name it and address cross-contamination after cooking. A later retail repack would require label and instruction review. This keeps supplier evidence, importer verification and the final manufacturer's controls in their correct places.

Measure appearance, insects and foreign material

Appearance is a contract subject with useful official vocabulary. USDA AMS frozen-broccoli standards define florets in §52.632(b) as heads with a portion of stem, 2.5–9 cm long and more than 6 g each. These are voluntary grading provisions unless incorporated into the purchase contract. For short spears or florets, Table VI expresses Grade A AQLs of 20.0 total, 6.5 major, 0.65 severe and 0.25 critical defects per 100 units; Grade B uses 25.0, 10.0, 2.5 and 0.65. The AMS floret tally sheet uses a 50-unit sample unit. Section 52.641 separately requires the applicable inspection sampling plan to determine lot sampling and acceptance. A 50-unit tally is not a universal lot acceptance sample, and an AQL value is not a promised percentage in every lot.

The style matters before using any number. If the actual pack contains small cut pieces instead of heads meeting the floret definition, Table VII for cuts and pieces uses a different 250 g basis divided into fifty 5 g increments and different AQL values. Chopped broccoli uses 2.5 g increments. If a customer orders florets but the lot includes detached buds, large stems or undersized heads, both sides should agree in the contract how those units are classified and whether the USDA floret style still fits. We can propose additional commercial targets for yellowing, broken pieces, woodiness, ice clumps or large stems, but those must be written as negotiated limits with sampling and measurement rules, kept separate from FDA law. A receiving complaint is much easier to resolve when both labs used the same thawing, sample selection and defect classification.

Illustrative scene: Broccoli florets arranged with a ruler for size and appearance grading.

FDA has a separate Compliance Policy Guide for frozen broccoli insects. Its enforcement action criterion is an average of 60 or more aphids, thrips or mites in any combination per 100 g. FDA's Food Defect Levels Handbook separately identifies AOAC 945.82 as the method. This is an action level for enforcement judgment, not a wholesome commercial allowance, a supplier target or permission to accept obvious infestation. We would set a stricter visual and sampling response in the buyer specification as appropriate, reject affected source material, investigate field and wash controls, and never advertise that "59 insects per 100 g is acceptable." An optical sorter may help with visible insects, but it cannot guarantee absence of small insects in dense florets.

Foreign material requires its own plan. Stone, metal, hard plastic, glass, soil, insects and unrelated plant matter have different detection routes and risks. We would establish incoming field inspection, sieves or air separation where suitable, visual sorting, magnet/metal detection or X-ray where actually installed and validated, and a final sample check. An X-ray check is based on detectable density contrast; a test card validates a specific material and size under specific product settings, not all materials. A metal detector has a different capability. If the buyer wants a specific detection sensitivity or zero visible foreign material contract, record the test pieces, challenge frequency, reject verification and lot disposition. A pretty photograph of a clean floret is not proof of a detector challenge.

Use an evidence-based lot release decision

Our proposed lot file begins with a signed specification and a supplier approval record. It then holds the grower lot, pesticide use, raw acceptance, production date and line, blanch/cool records, sanitation and foreign-material checks, packing count, temperature log, retained sample and any agreed laboratory certificates. The customer's FSVP importer may request additional supplier evaluation, audits, test results or corrective-action evidence. We can prepare those records for review to the extent the actual manufacturer can substantiate them; we should not substitute a website claim about "strict quality control" for a dated lot record.

Before production, GreenLand and the buyer sign an acceptance sheet stating the chosen size-measurement method, size band, stem rule, broken-piece and yellowing definitions, pack net weight, sampling units and random-draw method. The laboratory annex names each pesticide, legal residue definition and reporting limit, plus each agreed microbiological organism, method, sample mass, number of units and acceptance rule. The quality-release signatory is named on that sheet. A requested Grade A contract follows its stated USDA inspection plan; customer limits and the FDA insect enforcement criterion remain separate decisions.

Release has distinct decisions. First, do raw-material records and test results meet applicable U.S. pesticide tolerances and the buyer's narrower contract? Second, are the production and sanitation deviations closed with evidence? Third, does finished-product sampling meet the agreed appearance, foreign matter and microbiological criteria? Fourth, do packaging, label and cold-chain records identify the exact lot and cook-required use? Fifth, has the importer checked the FDA facility record, FSVP role, APHIS condition and tariff classification for the shipment? A failed first or second gate cannot be cured by giving the buyer a polished COA. We would hold material and agree disposition with the responsible technical and commercial parties rather than load while "waiting for paperwork."

GreenLand frozen broccoli product photograph for lot-form comparison.

Testing should be risk based. A field-pesticide screen selected from actual application and regional risk is more useful than a generic short panel. Multi-residue methods need adequate LOQs for each U.S. tolerance and must capture the legal residue definition; if the method cannot distinguish required epimers or metabolites, the report needs a different method or a qualified interpretation. Sample collection should cover the lot, not one chosen carton, and chain of custody should record who drew and sealed samples. An independent lab may be appropriate for a disputed result; a second test must address representativeness, not serve as a vote to ignore an initial failure.

FDA import alerts add a shipment check beyond our own COA. The actual manufacturer and product should be searched against current FDA import alerts before booking and again if circumstances change. A relevant listing can lead to detention without physical examination; it is not resolved by attaching a certificate with a vague "FDA approved" phrase. For broccoli, FDA's pesticide and filth concerns must be considered alongside the U.S. importer's evaluation of this supplier and the lot. If a hold occurs, preserve the lot, records and temperature condition while the importer handles the FDA notice and response deadline.

Pack and document a refrigerated U.S. shipment

For the quoted industrial pack, we would confirm carton net weight and liner material with the customer, check that a pallet can be sampled without damaging the rest, and carry the production lot on carton and pallet records. The line's food-contact material and seal design need to fit the frozen product and intended storage. We propose a pre-load count of cartons, net and gross weights, lot distribution, pallet map, container identity, seal number, set point and initial cargo temperature. A digital logger is useful when the buyer needs route evidence, but its position and review rule should be agreed in advance. A door opening can create a short air-temperature spike without proving product core thawed; an actual dispute should examine the logger trace, product condition and handling record together.

The export document package should match the food. Invoice and packing list descriptions should say single-ingredient blanched IQF broccoli, China origin and the actual floret form, not an abbreviated "veg mix." Manufacturing facility identity, lot and carton count must align with the food registration, FDA prior notice and customs entry. The U.S. buyer appoints its FSVP importer and broker; the ocean ISF party files in advance; the manufacturer maintains its relevant FDA registration and U.S. agent. Our U.S. import guide sets out those responsibilities and the 2026 renewal, APHIS, prior notice and release path in order. These are different obligations from an optional USDA Grade A contract.

Illustrative scene: Frozen broccoli packed in a lined bulk carton.

Certification requests are assessed by trigger and holder. FDA food facility registration is a regulatory facility step, not a food certificate. A customer may require a GFSI-recognized food-safety audit, HACCP-based program or third-party site certificate under its supplier policy, but we should provide only a current certificate for the actual producing site and scope after verification. U.S. organic marketing requires the relevant USDA National Organic Program certification and import certificate/traceability route for organic goods; it cannot be inferred from a non-detect pesticide test. Kosher and Halal are customer, channel or religious-program requirements with their own certifying bodies and product/site scopes, not general U.S. legal entry conditions for conventional broccoli. None is claimed as already held here.

The first commercial trial should define the buyer's cooking application, floret and stem dimensions, defect sampling basis, agreed lab panel and LOQs, packaging weight, destination port, volume and desired delivery window. We can return a proposed spec and evidence checklist for the named source facility, then confirm what can be produced and documented before an order is signed. The shipment should move only after the U.S. applicant holds the APHIS PPQ Form 587 permit required by the current ACIR frozen-Brassicaceae record, the importer confirms its quick-freeze and Part 305 handling conditions, the broker validates HTS 0710.80.97.25 for the actual non-organic bulk form and current China measures, and the respective FDA and customs filing parties accept their roles. That is a workable supply conversation grounded in records rather than a blanket claim of compliance.

For the trial order, our document coordinator maps each finished carton and pallet back to the source and production lots, including any splits or commingling. Before loading, the named quality reviewer signs the laboratory, process, pack and identity checks, and the export coordinator confirms the issued permit and agreed document version with the importer. Unresolved safety, pesticide or identity deviations keep the lot on hold. At arrival, the buyer returns seal and carton counts, receiving product temperatures, logger records, damage or defect findings and the FDA/CBP release references against the same lot IDs. A claim is investigated against that defined material before use or resale.