A field-tested playbook for verifying true net weight and glaze on Indonesian frozen seafood during PSI in 2026. Includes tools, AQL-based sampling for a 20’ reefer, step-by-step deglaze to avoid drip loss, on-site calculations, acceptance tolerances, and reporting tips buyers actually trust.
If you’ve ever stood in a cold room arguing over “short weight,” you know why a clean net weight and glaze check is the difference between a smooth release and a costly claim. We’ve run hundreds of PSIs across Indonesian plants, and the same truth keeps showing up. Your method matters more than your opinion. Here’s the exact, field-tested approach we use in 2026 for Indonesian frozen seafood.
The 3 pillars of a defensible PSI net weight program
-
Fit-for-purpose tools with traceable calibration. A great procedure won’t save you if the scale drifts. We verify with certified weights before any measurement and log the readings.
-
Sampling that truly represents the reefer, not just the easy-to-reach pallets. We randomize, span front/middle/rear, and cover multiple production lots when present.
-
A deglaze procedure that removes ice without warming the core. That’s how you avoid drip loss and false “short weight.” We standardize water temperature, time, and blotting.
This framework applies whether you’re checking Frozen Shrimp (Black Tiger, Vannamei & Wild Caught) in 1 kg bags or IQF fillets like Grouper Fillet (IQF) and Mahi Mahi Fillet.
Week 1–2: Prep and validation (tools, plan, and specs)
-
Tools we bring on-site:
- Scales: 5 kg bench scale with 1 g readability for unit packs. 30 kg platform scale with 5 g readability for cartons. Both Class III or OIML R76 compliant.
- Certified test weights: 500 g and 2 kg with current certificates.
- Thermometer: probe or IR, range −20 to +50 °C.
- Deglazing kit: stainless sieve/colander (5–8 mm holes), chilled potable water, stopwatch, untreated paper towels, tongs, gloves, clean tray.
- Documentation: printed inspection record sheet, pre-numbered sample labels, camera/phone for photos.
-
Calibrated scale verification:
- Level the scale. Warm up for 10 minutes.
- Zero test: pan empty reads 0.0 g.
- Span test: place 500 g and 2,000 g weights. Record displayed values. Allowable error should be within the scale’s stated tolerance. If not, re-calibrate or replace.
-
Align the spec in writing. In our experience, most 2026 buyer specs require:
- Label net weight refers to product only, excluding glaze.
- Glaze target range, often 6–10% of product weight for IQF fillets and shrimp.
- Acceptance: lot average must be ≥ 100% of declared net weight. Individual packs not below 97–98% for retail, 95–97% for foodservice, unless otherwise specified. Agree the exact tolerance upfront.
Tip you’ll use forever: pre-chill your sieve and trays. Warm metal accelerates melt and can inflate “drip loss,” especially on thin fillets like Sweetlip Fillet (IQF).
Week 3–6: Execute the PSI (sampling and the deglaze that works)
What sample size should I use in a 20’ reefer of shrimp?
For a 20’ reefer with ~1,000–1,200 cartons of 1 kg shrimp bags, here’s a practical plan that buyers accept and inspectors can execute:
- Carton selection: randomize 10 cartons from at least 5 different pallets. Cover front/middle/rear of the container or cold store.
- Unit selection: test 3 inner bags per selected carton. Total 30 bags measured.
- If the lot mixes production dates or batches, split the sample proportionally and report results by batch.
This sits between ANSI/ASQ Z1.4-style lot coverage and the realities of time and cold-chain. For high-risk or claim-prone SKUs, we go to 48–60 units.
How do I measure glaze percentage during PSI?
You’ll record both net weight and glaze. Two useful expressions exist. Decide which one your buyer wants.
- Inputs: Wg = gross frozen weight with glaze. Wp = product-only weight after deglaze.
- Glaze weight = Wg − Wp.
- Glaze % of product weight (most common in seafood specs): (Wg − Wp) / Wp × 100.
- Glaze % of total weight (used in some QA teams): (Wg − Wp) / Wg × 100.
We always write the formula used right on the sheet and in the report. It ends arguments later.
Step-by-step deglaze test for frozen shrimp and IQF fillets
-
Weigh frozen unit immediately out of carton. Record Wg. Photo the reading with the pack visible.
-
Deglaze under a gentle stream of potable water at 10–15 °C. Rotate the pack contents in a chilled sieve. Keep the core frozen. Typical time: 20–60 seconds for shrimp, 30–90 seconds for fillets.
-
Stop when the surface looks matte, not glossy, and pieces no longer feel slippery. Don’t keep washing. You’re removing ice, not thawing product.
-
Drain on the sieve for 30 seconds. No squeezing. Then quick blot with a single layer of paper towel to remove visible surface water.
-
Weigh immediately. Record Wp. Photo the reading again.
-
If the product starts to soften, pause testing, return samples to −18 °C for 15–20 minutes, then resume.
That’s the entire trick. Short, cold, and timed. I’ve found even a 30-second over-wash can drop 1–2% on delicate items.
How long should I thaw samples before weighing?
You don’t thaw. You only deglaze to cold product. Thawing creates drip loss and falsely low net weights. If you must check “drained weight” for a cooked or brined spec, that’s a different method and should be specified separately.
Can I check net weight without fully deglazing?
For a quick screen on the floor, yes. Two options:
- Glaze thickness check on fillets: scrape a corner and measure ice thickness with calipers. 0.5–1.5 mm often corresponds to 6–10% glaze on medium fillets. Use as a flag, not a final.
- Partial spray wash: 10–15 seconds to remove obvious glaze, then weigh. If results are close to label, proceed to full deglaze on a subset. For acceptance, always rely on full deglaze results.
On-site calculation and acceptance example
Spec: 1,000 g net shrimp, glaze 6–10% of product weight.
- Unit 1: Wg = 1,096 g. After deglaze Wp = 1,000 g. Glaze% = (96 / 1,000) × 100 = 9.6%. Pass.
- Unit 2: Wg = 1,070 g. Wp = 980 g. Glaze% = 9.2%. Net short by 20 g (−2.0%). Depending on tolerance, may be acceptable as an individual, but watch the average.
- After 30 units: Average Wp = 1,002 g. No unit below 970 g (−3%). If your tolerance is average ≥ 100% and no unit below −3%, this lot passes. If your buyer set −2% minimum per unit, it fails because of the 980 g unit.
We verify carton average vs. individual units and state the rule used. That clarity prevents disputes.
Week 7–12: Scale and optimize across suppliers
- Standardize the record sheet. Include product ID, species, form (e.g., Red Snapper Portion (WGGS / Fillet)), production date, declared net weight, target glaze, scale ID, calibration check, water temperature, Wg/Wp per unit, glaze %, mean, min, max, and acceptance statement.
- Photographic evidence. We capture: scale zero, Wg and Wp screens, pack label, randomization map, and deglaze setup. Buyers trust what they can see.
- Spot-train factory QA. Shared technique reduces variance. Many short-weight disputes disappear once everyone follows the same stopwatch.
What’s interesting is how quickly this tight method reduces claims. We’ve seen repeat lots go from 20% rework to near zero in two shipments.
Quick answers to the questions we get most
What tools do inspectors need to check net weight and glaze on-site?
- Class III/OIML-compliant scales with 1–5 g resolution.
- Certified weights (500 g, 2 kg).
- Chilled potable water supply and thermometer.
- Stainless sieve/colander, stopwatch, paper towels, tongs, gloves.
- Pre-chilled trays and a camera.
What’s an acceptable net weight tolerance in 2026?
Most buyers require lot average ≥ declared net weight. Individual packs typically not below 97–98% for retail and 95–97% for foodservice. Some markets apply NIST Handbook 133 or EU Average Quantity rules. Bottom line. Put the exact tolerance in the purchase spec and reference it in the PSI report.
How should I document results for buyers or customs?
Use a one-page summary plus appendices:
- Summary: product, lot, sample size, mean net, min/max, glaze mean, acceptance statement.
- Appendices: raw data table, photos, scale calibration check, sampling map, signatures.
A neat, transparent pack of evidence gets shipments cleared faster.
The 5 biggest mistakes that kill net weight inspections
- Warm water and long washes. You’re cooking the test. Keep water 10–15 °C and keep it short.
- No scale check. If you didn’t verify with certified weights, your numbers are debatable.
- Convenience sampling. Only front pallets, only top layers. Randomize across the lot.
- Blotting forever. One light blot. Excessive blotting removes surface moisture beyond glaze.
- Mixing formulas. Glaze % of product vs. of total weight. Declare which one you use.
Avoid those, and your acceptance calls will hold up under scrutiny.
Practical templates you can copy
-
Net weight and glaze worksheet (fields):
- Product/Lot/Date/Inspector
- Declared net weight and target glaze
- Scale ID and calibration check results
- Water temp and room temp at test time
- For each unit: Wg, Wp, Glaze%, Remarks, Photo ID
- Summary: Mean Wp, Min/Max, Mean Glaze%, Acceptance, Sign-off
-
Sampling map for a 20’ reefer: pallets 1–10. Select pallets 1, 3, 5, 8, 10. From each, pull cartons from top, middle, bottom positions. Label positions on the map and photos.
Where this advice applies (and where it doesn’t)
- Applies to: IQF shrimp, fillets, steaks, portions, WGGS with glaze. Think Kingfish Fillet (Portion Cut / IQF) or Mahi Mahi Portion (IQF).
- Doesn’t cover: microbiological or chemical testing, permits, HS codes, or logistics. Different methods apply to brined, cooked, or soaked products. Align with the buyer’s spec before testing.
Recent trend we’re seeing. Buyers are tightening glaze bands and asking for photo evidence of deglaze steps, not just outcomes. We now include short video clips for claim-prone SKUs like sashimi items such as Yellowfin Saku (Sushi Grade) where over-washing can affect color.
Need a lot-specific PSI checklist or help tuning tolerances to your market? You can Contact us on email. If you’re on the floor right now and need a quick ruling, Contact us on whatsapp.
The reality is simple. A repeatable method protects everyone. Do the boring basics right, and your net weight and glaze checks will stand up from the factory floor in Indonesia to the buyer’s dock halfway around the world.