Factory planning for water bottling projects
Legal metrology · Fill-process control
Net-content control has two simultaneous aims: comply with the declared quantity rules that apply in the destination market and run the filler with no more unnecessary overfill than process variation requires. The team must first identify the legal metrology basis, reference conditions, sampling interpretation and marking obligations for each market. It then needs a capable mass or volume method, traceable standards, density or temperature treatment where relevant, and data that can be mapped to individual filler heads and operating states. NIST Handbook 133 and European Commission prepackage resources illustrate different official frameworks; neither should be copied worldwide. No universal tolerance or sample plan belongs on this page because the competent authority and product route define them.
01
Quality, operations and regulatory owners should maintain a market register for each declaration and pack. Capture whether evaluation is by mass, volume or another recognized basis; how individual and average requirements interact; which reference condition applies; and how official sampling differs from routine process control. A generic target above the label is not a compliance method.
Link each SKU and destination to the current competent-authority rule, permitted units, nominal quantity marking and evaluation procedure. Keep the legal source and review date. Do not assume that passing one market's average test satisfies another market's individual-package provisions.
State how product temperature, density, dissolved gas, evaporation or meniscus interpretation affects the result where relevant. Use validated calculations and representative product data. A volumetric setpoint observed warm may not equal the declared volume under the official basis.
Compare gravimetric, volumetric, checkweigher or other approved methods for resolution, range, repeatability, speed and operator influence. Demonstrate that uncertainty is small enough for the decision; more decimal places on a display do not improve measurement capability.
Define who approves the method, daily control limits, adjustment, product hold, investigation and market release. Operators may make bounded process corrections, while legal disposition of potentially nonconforming lots should remain with an authorized quality or regulatory role.
02
Routine checks should reveal both central tendency and variation by filler position, time and event. Design the sampling interval from line speed, failure detection, process stability and applicable rules, then preserve each result with head or valve identity where technically possible. Pooling all bottles into one daily average can conceal an underfilling position.
Use suitable, identified mass or volume standards and verify balance, reference vessel or checkweigher status. Establish bottle and cap tare variation where a gravimetric conversion requires it. Never subtract a convenient historic tare from a materially changed package without study.
Preserve head, valve or carousel position, time, format, recipe and operating state. Use a controlled coding or collection method and verify that position numbering aligns with maintenance records. This converts a compliance result into actionable equipment evidence.
Include startup, speed change, tank-level variation, after adjustment, changeover, stop-restart and end of run as the risk analysis requires. A stable mid-run sample cannot qualify transient product automatically.
Document repeatability, reproducibility, resolution, environmental effects, density conversion and sampling variability. Use this evidence when setting internal action limits so the process does not oscillate around measurement noise or run dangerously close to a legal boundary.
| Control layer | Evidence captured | Primary decision | Responsible role |
|---|---|---|---|
| Market master | Current rule, nominal declaration and reference basis | Which official evaluation applies? | Regulatory or legal-metrology owner |
| Measurement system | Method study, standards, tare/density and checks | Can the result support disposition? | Quality laboratory or metrology |
| Process sample | Value, time, head, recipe and event state | Adjust, continue, investigate or hold | Trained operations with quality rules |
| Lot review | Distribution, exceptions and affected interval | Release or controlled disposition | Authorized quality owner |
03
Giveaway is often the margin used to cover uncontrolled variation. Lowering the global setpoint before understanding spread can create underfilled bottles. Separate common variation from head-specific, package-specific and event-related causes, correct the source, then make bounded setting changes with heightened sampling and an approved rollback rule.
Review mean, spread, tails and time sequence using the appropriate statistical and legal interpretation. Compare by head, shift, speed and format. Do not select a process capability index or target without confirming distribution assumptions and the actual authority-defined specifications.
Investigate valve wear, venting, timing, product pressure, bottle height or neck variation, foaming, nozzle condition and carousel mechanics. Use repeated position evidence and maintenance inspection. Recalibrating the whole filler can hide one failing valve while increasing giveaway elsewhere.
Assess temperature, density, aeration, water level, bottle internal volume and deformation under line conditions. Link package changes to method and setting review. A lighter or differently shaped bottle can alter presentation and apparent fill even at the same nominal capacity.
Define the intended change, affected format, starting state, sample intensity, stop criteria and approver. Confirm multiple carousel revolutions or other representative coverage and the legal basis before adopting the new center. Preserve before-and-after raw data.
04
A failed sample must trigger a known boundary and decision, not an improvised extra bottle. Use the last verified acceptable state, filler residence and accumulation to define potentially affected product. Confirm measurement validity promptly, but keep product controlled while process and compliance implications are evaluated.
Trace time, head, pallet and lot from the last acceptable check through detection and correction, including bottles in conveyors or rework. Use a conservative boundary when mapping is uncertain. A later good sample does not erase earlier uncontrolled production.
Verify standard, balance or checkweigher state, tare or density data, sample identity and method execution, then examine filler conditions. Distinguish a measurement-system failure from real underfill; both require documented correction before release.
Apply the relevant official evaluation and the site's approved plan to the affected quantity. Rework, relabel or other options require technical and legal authorization. Do not average unrelated lots to make a failing group appear acceptable.
Assess new bottle, cap, product profile, filler part, software, speed or measurement method before routine use. Trend giveaway cost and compliance signals together, but never let cost pressure override the authority-defined release decision.
R
These sources support the risk-control method on this page. They do not set project-specific legal limits, test frequencies, engineering values or approvals; verify the current edition and local applicability before a decision.
U.S. NIST
European Commission
Codex Alimentarius
Buyer questions
It depends on the destination market, package and current official procedure. Confirm the competent-authority rule; do not apply a tolerance copied from another country or product.
The process center must account for legal requirements, measurement uncertainty and demonstrated variation. First reduce assignable variation, then set a controlled target supported by evidence.
It can provide powerful in-line control if properly specified and verified, but official or release methods may require additional evidence. Connect it to controlled standards, tare/density and product disposition.
Position mapping finds a low or unstable valve that a pooled average can hide. It also makes maintenance and post-adjustment verification more precise.
Provide the SKU/market register, current legal basis, bottle and cap data, product density or temperature behavior, filler design, measurement method study, position-level samples and hold procedure.
Move this project question forward
Giveaway is often the margin used to cover uncontrolled variation. Lowering the global setpoint before understanding spread can create underfilled bottles. Separate common variation from head-specific, package-specific and event-related causes, correct the source, then make bounded setting changes with heightened sampling and an approved rollback rule.
Not sure which data matters? Send what you have and state the decision you need to make.
Send target capacity and SKUs, source-water report, site utility schedule, building layout and required project milestones.
The project desk can identify missing inputs and a practical next step. Final engineering, configuration, compliance and commercial terms remain project-specific.
Allot Tech (Suzhou) Co., Ltd. · sales@allottech.com · Project telephone: +86 186 6213 1120 / +1 818 262 0958