How do I pick the risk threshold for guardbanding?

Choose a consumer-risk alpha aligned to regulation and brand risk. Validate against historical distributions and worst-case shifts before deploying.

Can I reduce guardband without increasing sub-TNE risk?

Yes. Reduce total variability and measurement uncertainty (process stabilization, tare governance, device bias reduction) so the same risk is achieved with a smaller offset.

What if checkweigher and lab balance disagree?

Treat it as a data integrity signal. Verify calibrations, re-sample, and reconcile. If bias persists, pause release and document under Deviation/CAPA.

Do I need to recompute guardband every packaging lot?

Recompute when total sigma changes materially (new packaging lot variability, device changes, formulation shifts) or at defined review intervals.

Tolerable Negative Error (TNE)Glossary

Tolerable Negative Error (TNE) – SPC Guardbanding

Q: Is TNE the same as my internal tolerance?

No. TNE is a legal threshold for individual units; internal tolerances trigger operational interventions. Set internal limits tighter than TNE to maintain compliance margin.

Q: We claim volume but weigh mass—what changes?

Treat density and temperature correction as measurement uncertainty. Version and verify these factors and include them in the total sigma used to set guardband.

This topic is part of the SG Systems Global regulatory & operations glossary.

Updated October 2025 • Weights & Measures, Net Content Control, Risk‑Based SPC • Packaging, QA, Manufacturing

Tolerable Negative Error (TNE) is the legally permitted shortfall below a product’s nominal declared quantity for individual packages under average‑system net content rules. In practice, regulators allow a small, bounded negative deviation for some units in a batch provided the batch average meets the label claim and very few (or none) fall beyond the TNE threshold. Operationally, this converts to a statistical problem: set the process target and control strategy so the probability of producing sub‑TNE units is acceptably low while avoiding excessive overfill. That strategy is known as SPC guardbanding—a disciplined offset above nominal informed by SPC, measurement uncertainty, tare variability, and legal tests for net content.

“Guardbanding pays twice: it protects consumers from under‑fill and protects manufacturers from chronic over‑give.”

TL;DR: TNE is the legal shortfall allowed for some units. Build a risk‑based guardband by combining your process variation, control limits, and total measurement uncertainty (scale + weighing + label/conversion logic) so the probability a unit falls below nominal‑TNE is below your chosen risk threshold, while the batch mean remains on or slightly above nominal.

1) What TNE Guardbanding Covers—and What It Does Not

Covers: the design of net‑content targets and SPC rules at fill/pack. It governs how to set the filling mean above nominal (the guardband), how to monitor drift and variability using control charts, how to incorporate laboratory cross‑checks and checkweigher streams, and how to document conformance to statutory tests. TNE guardbanding also covers recipe logic for density conversions (when claiming volume but controlling by mass), tare handling, and capability improvement.

Does not cover: redefining legal limits or substituting SPC for regulation. TNE values, sampling schemes, and acceptance rules are set by jurisdiction. Your system must demonstrate compliance; it cannot redefine compliance.

2) Legal, System, and Data Integrity Anchors

TNE arises from weights‑and‑measures law (e.g., average‑system net content). Electronic evidence supporting TNE guardbanding must meet 21 CFR Part 11/Annex 11: validated software under CSV, unique users, and immutable audit trails. Devices must be qualified (IQ/OQ/PQ) and in active calibration status. Tare values and density factors sit in version‑controlled master data governed by Document Control.

3) The Evidence Pack for TNE Compliance

A complete “TNE pack” shows: the governing legal basis and applicable TNE for the SKU; the guardband rationale and calculation; current process capability (σ, Cp/Cpk); combined measurement uncertainty (scale, checkweigher, tare, density); control plan (chart types, sampling frequencies, checks); empty‑pack verification and tare logic; and batch records proving the applied target, observed means/variances, and disposition of any nonconformances under Deviation/CAPA. Where volume is claimed but mass is measured, include density traceability and temperature corrections.

4) From Goods Receipt to Label Claim—A Standard Path

1) Goods Receipt & Component Control. Packaging enters via Goods Receipt; QA verifies tare by lot and performs empty‑pack sampling before Component Release.
2) Line Setup in MES. The MES loads the current guardband target and control limits; scales and checkweighers are verified in status.
3) Execution & SPC. Gross, tare, and computed net values feed control charts; on trend/rule breaks, operators adjust fill or stop per the control plan.
4) Release. Batch means and individual distributions are assessed against legal tests; labels pull the same computed net used for disposition; inventory moves to shippable in the WMS after QA review.

If any prerequisite fails (device out of status, wrong tare set, density factor expired), block execution and document remediation before proceeding.

5) Designing the Guardband—A Practical Method

Define the guardband as the offset between nominal and the process mean target. One practical approach is: (i) identify the legal TNE for the nominal; (ii) estimate total standard deviation σ_tot combining process variation and measurement uncertainty; (iii) choose a consumer‑risk threshold (e.g., the probability a unit is below nominal − TNE must be < α); (iv) set the target so the lower tail meets that criterion (for normal processes, target ≈ nominal + TNE + z_α·σ_tot − k, where k reflects the acceptance rule’s allowance for some sub‑TNE units). Validate with historical data and challenge with worst‑case shifts (density drift, tare change, warm‑up effects). Re‑estimate guardband when σ or uncertainty changes materially.

6) Measurement Uncertainty & Tare Variability

Under‑fills are often created by small uncertainties that add up. Combine: scale repeatability and bias; checkweigher bias vs. reference balance; tare variability by packaging lot; density and temperature corrections; and rounding/printing rules. Treat uncertainty components as independent unless evidence suggests correlation; combine via root‑sum‑of‑squares to obtain σ_tot. If you cannot quantify a component, bound it conservatively until studies provide better estimates. Governance for tare and density belongs under Document Control with periodic verification.

7) Data Integrity—Proving Compliance Without Recalculation

Every net value should be reconstructable from stored gross, applied tare (by lot), device IDs, and recipe logic. The audit trail must show the guardband version used and any target adjustments, with electronic signatures. Avoid spreadsheet shadow systems; implement guardband and control limits in validated MES/WMS integrations under CSV.

8) Sampling, Checkweighers & Lab Cross‑Checks

In‑line checkweighers provide fast screening; labs provide traceable anchors. Define a plan that compares checkweigher means to a calibrated reference balance at set intervals. If sustained bias appears, treat it as a signal: halt release, re‑establish status, and reassess guardband. Tie calculations and methods to Laboratory Analyses & Review.

9) Equipment Status—Minimum Weight, Warm‑Up, and Drift

Set your control plan around device physics: respect minimum weight, warm‑up time, zero tracking, and vibration limits. If a device goes out of status during a batch, perform impact assessment on units potentially below TNE and manage via Deviation/CAPA before release. Record device IDs with each net calculation to preserve traceability.

10) Labels, Claims & Conversion Logic

Guardbanding only protects you if label claims reflect the same computation used for release. Labels should pull computed net from the execution record, not re‑calculate. If you claim volume but fill by mass, density and temperature factors must be versioned and audit‑traced. Validate label content with Label Verification and keep density/tare effective dating under Document Control.

11) Warehouse Status & Logistics Units

Cases and pallets introduce secondary tare. Ensure the WMS uses validated carton/pallet tare and that inventory remains in Quarantine/Hold until QA verifies TNE/guardband compliance. Pack & Ship then proceeds using FEFO/FIFO rules with accurate weights downstream.

12) SPC in Daily Control—Keeping Risk Low and Waste Lower

Chart individual nets or subgroup means with alert/action limits. Track center and spread, apply runs/trends rules, and use Cp/Cpk to quantify headroom to TNE. Where processes creep low, bias the target up temporarily and launch root‑cause work on the drivers (nozzle wear, viscosity, temperature control, tare variability). The goal is a small guardband and a stable process—not permanent over‑give.

13) Metrics That Demonstrate Control

Risk of Sub‑TNE (estimated % units below nominal−TNE) by SKU and line.
Over‑give Cost (mean − nominal) × volume, trended with guardband changes.
Cp/Cpk of Net Weight and σ_tot components (process vs. measurement).
Checkweigher Bias vs. Lab and time‑to‑correction for significant bias.
Tare Verification Pass Rate by packaging lot/supplier.
Label/Record Mismatch Incidents (should be zero).

These KPIs connect legal risk (sub‑TNE probability) to financial impact (over‑give) and operational levers (σ and bias).

14) Common Pitfalls & How to Avoid Them

Assuming σ from last year still holds. Re‑estimate σ_tot when materials, equipment, or environment change.
Ignoring measurement uncertainty. Include tare, density, and device bias; do not rely on process σ alone.
Shadow spreadsheets for guardband. Implement targets/limits in validated systems with audit trails.
Label/net disconnect. Force labels to consume the executed net; validate with automated scans.
One‑size‑fits‑all guardband. Different SKUs, viscosities, and packs need different offsets.
Late discovery at final QA. Use in‑process SPC and automated interlocks to prevent accumulations of sub‑TNE units.

15) What Belongs in the TNE/Guardband Record

Include the SKU and nominal declaration, applicable legal TNE, guardband methodology and current offset, σ_tot breakdown, control plan, tare/density master data references, device IDs and status, chart summaries, out‑of‑control events with dispositions, and approvals/effective dates under Document Control. Retain per record retention rules aligned to product life and jurisdictional requirements.

16) How This Fits with V5 by SG Systems Global

Guardband as Master Data. In the V5 platform, each SKU can carry a regulated TNE attribute and a calculated guardband offset. These are versioned, effective‑dated values under Document Control, with change reasons and approvals captured in the audit trail. When packaging lots are scanned, V5 also applies the correct tare table and density factors by lot to ensure the executed net aligns with the declared claim.

Execution & Interlocks. The V5 MES pushes the current target and control limits to scales and checkweighers. If a device is out of calibration status, the lot is wrong, or the observed distribution raises sub‑TNE risk, execution blocks and opens a guided workflow (investigation, adjustment, re‑verification) before continuation.

SPC & Analytics. V5 streams individual net readings into live SPC with rules for runs, trends, and point‑beyond‑limits. Dashboards estimate sub‑TNE probability in real time and quantify over‑give cost. Drill‑through shows the contribution of process variability vs. measurement uncertainty, helping you target the cheapest improvement first.

Label & WMS Alignment. V5’s label engine reads the same executed net used for release, eliminating re‑calculation drift. The V5 WMS enforces Quarantine/Hold until QA confirms guardband/TNE compliance, then supports compliant Pack & Ship.

Bottom line: V5 operationalizes TNE guardbanding—targets, devices, tare/density, SPC, labels, and release are all synchronized so legal risk drops and over‑give shrinks.

17) FAQ

Q1. Is TNE the same as my internal tolerance?
No. TNE is a legal threshold for individual units; internal tolerances define when you intervene operationally. Good practice sets internal limits tighter than TNE to ensure compliance margin.

Q2. How do I choose the risk threshold for guardbanding?
Pick a consumer‑risk α that aligns with regulation and brand risk (commonly single‑digit ppm to low tenths of a percent). Validate against historical distributions and worst‑case shifts.

Q3. We claim volume but weigh mass—what changes?
Treat density and temperature correction as part of measurement uncertainty. Version and verify these factors; include them in the σ_tot used to set guardband.

Q4. Can I reduce guardband without risking sub‑TNE?
Yes—by lowering σ_tot (stabilize the fill, improve tare governance, reduce device bias) you can keep risk constant with a smaller offset above nominal.

Q5. What if checkweigher and lab disagree?
Treat as a data integrity signal. Re‑check calibrations, re‑sample, and reconcile. Pause release if bias persists; document under Deviation/CAPA.

Q6. Do I have to re‑compute guardband for every packaging lot?
Not necessarily. Recalculate when σ_tot changes materially (new packaging lot with different tare variability, new device, new formulation) or at defined review intervals.