Ask a filling machine supplier what accuracy their machine achieves and you will get a number with a plus-or-minus sign in front of it. Ask the inspector in your destination market the same question and you will get a table, a sampling rule and a legal limit on how far below the declared quantity any single package may fall. Those are two different conversations, and treating the first as if it answered the second is how a shipment gets rejected at the border.
The cost is not only the seized batch. A line that is set up to hit the average and no more will fail intermittently, because the failures depend on the sample the inspector draws, not on the average you control. The instinctive fix — give away product by filling high — is the most expensive compliance strategy available, because the giveaway is paid on every single package and is invisible on the shift report.
Sailwin has built filling and packaging machinery for 15+ years, with 500+ machines delivered into 60+ countries, CE marking, ISO 9001:2015 manufacturing and a 2-year whole-machine warranty. The filling range covers three-in-one rinsing, filling and capping monoblocs, isobaric filling for carbonated product, hot filling, and 5-gallon lines, in SUS304 and SUS316L product contact materials, each subjected to a 24-hour factory acceptance run. This article explains what filling accuracy standards mean in practice, how accuracy differs from precision, how to verify both on your own line, and where drift actually comes from.
Key Takeaways
- Accuracy and precision are different targets. Accuracy is how close the average is to nominal; precision is how tightly the packages cluster. Legal metrology constrains both, and a machine can satisfy one while failing the other.
- The tolerance you must meet is a legal number, set by the destination market’s prepackaged goods regulation, not by the machine datasheet.
- Filling high is a cost, not a control strategy. Deliberate giveaway is paid on every package and is the most expensive way to avoid a compliance failure.
Send Your Product, Volume and Destination Market for an Accuracy Review
Share the bottle or can size, product type and target market — our engineers return the filling system, the valve arrangement and the verification routine that fits your tolerance.
1. Accuracy, Precision and the Legal Number You Actually Have to Meet
Accuracy is a statement about the mean. Precision is a statement about the spread. A filler can be accurate and imprecise: the average lands on nominal, but individual fills scatter widely, so some packages fall short and some overfill. It can be precise and inaccurate: every package is filled to exactly the same volume, and that volume is consistently below the declared quantity. Only the second failure is a legal problem, and only the first is a customer complaint.
Legislation for prepackaged goods handles both by combining an average rule with a shortfall rule. The average of a batch must be at least the declared nominal quantity. A restricted minority of individual packages may fall below nominal, but only within a tolerable negative error for that nominal size. No single package may fall below nominal by more than twice that tolerance. This structure is why filling to the nominal figure exactly is a losing strategy: a process centred on nominal produces roughly half its packages below nominal by definition.
The frameworks you will meet by market are the OIML R87 recommendation on prepackages, the European Measuring Instruments Directive for instruments and the corresponding prepackaged goods rules, and national handbooks such as those published by the United States weights and measures authorities. They are not identical, and the numeric limits depend on nominal quantity and on the product’s unit of measure. The practical point for a filling line is simpler than the legislation: target the process centre above nominal by a calculated margin, and treat the spread as the variable you have to control.
Work out that margin rather than guessing it. It depends on three things:
- The tolerance rule for your nominal size in the destination market, since the allowable shortfall scales with the declared quantity.
- The standard deviation of your filler at each valve, because the margin has to cover the tail of the distribution rather than the average.
- The measurement uncertainty of your own checkweighing, because a scale that is drifting will report compliance that does not exist.
Get a Fill Control Plan for Your Product and Package
Send the nominal volume, product density and destination market — we return the target fill, the process centre margin and the sampling routine to verify it on the line.
2. Where Filling Accuracy Drift Actually Comes From
Most accuracy investigations begin with the filling valve and end somewhere else entirely. The table below maps the symptoms operators report onto the causes that usually produce them.
| Symptom | Usual cause | Where to check first |
|---|---|---|
| One or two heads consistently out | Valve wear, a partially blocked nozzle, or a seating problem on that head only | Per-head weight data, not the line average |
| Drift through the shift | Product viscosity changing with temperature, or supply pressure falling as the tank empties | Product temperature log against fill weight |
| Foaming and short fills | Turbulence at the nozzle, fill speed too high for the product, or excessive headspace pressure | Fill valve flow profile and nozzle design |
| Drips and cap contamination | Nozzle cut-off timing, worn seals, or capping torque set without a measured target | Cut-off adjustment and capping torque calibration |
| Step change after cleaning | Residual cleaning fluid in the product path or a change in valve seating after reassembly | Line-layout and cleaning-cycle records |
| Everything reads high | Checkweigher calibration, tare assumption, or product density assumed rather than measured | Reference weights and a measured density figure |
The line average is the least useful number in a filling hall. Per-head data is what identifies a cause; the average only tells you whether the problem is currently visible to a customer.
3. How to Verify Filling Accuracy on Your Own Line
Verification is a sampling plan, not a spot check. Take a sample at the start of the run, again at the middle and again at the end, because the whole point is to detect drift that a single sample cannot show. Weigh the packages rather than trusting the machine’s own fill-volume display, and weigh them on a scale that has a current calibration certificate and a documented tare procedure.
Attribute the samples to heads. If the machine fills 24 or 36 heads, an unattributed sample produces a line average that hides everything worth knowing, because one worn valve among thirty-six will disappear into the noise. Most filling machines can be indexed so that consecutive packages map to known heads, and a control chart per head will find a failing valve long before a customer does.
Finally, convert weights to volumes using a measured density, not a nominal one. On a hot fill line this matters twice over, because density changes with temperature, and a measurement taken on a cooled sample has to be interpreted against the fill temperature. Sailwin monoblocs hold product temperature control to about ±1 °C through PID control, which reduces one source of variation, but it does not remove the need to state the temperature at which the sample was taken.
4. Matching the Filling System to the Tolerance You Need
Different filling principles produce different spread characteristics, and choosing between them on price alone usually costs more in product giveaway than the machine difference. Gravity and laminar-flow filling suit low-viscosity product and are gentle on the container, but they are sensitive to supply pressure and headspace conditions. Piston fillers meter a fixed volume mechanically and are relatively indifferent to viscosity, which helps with sauces and pulps, but they need careful seal maintenance. Flowmeter filling closes the loop with a measurement rather than a fixed displacement, which makes it tolerant of product change but dependent on correct meter calibration. Isobaric filling is the only route for carbonated product, where the container is pressurised to the product’s equilibrium pressure before and during filling.
Sailwin’s filling range covers the three-in-one monobloc arrangement in which rinsing, filling and capping happen in one frame, a 180° bottle turnover for rinsing, laminar flow filling valves, isobaric filling for carbonated soft drinks, hot filling for juices and 5-gallon line configurations, with product contact parts in SUS304 and SUS316L. The monobloc layout reduces bottle handling between stages, which also removes a source of contamination and of fill variation. Constant magnetic torque capping protects the neck finish from over-torque, and 24-hour factory acceptance testing gives the accuracy figure a defined basis before the machine ships.
5. Case Study: Bringing a Hot Fill Line Back Under Control
A juice producer was filling hot product to a deliberately high target to clear a compliance audit. Product giveaway had become a fixed cost of doing business.
- Fill target raised to a level that cleared checks but gave away product on every package
- Line average used for decisions, so a small number of underfilling heads stayed invisible
- Fill weight drifting through the shift as product temperature moved in the supply tank
- Samples re-indexed to individual filling heads so per-head weight data replaced the line average
- Three-point sampling introduced across each run, with product temperature recorded at the same time
- Fill target recalculated from the measured spread and the applicable shortfall tolerance rather than from habit
- Nozzle cut-off and capping torque re-established against a measured target
- The spread was controlled rather than the average, which is what allowed the target to be lowered at all
- Giveaway was converted into a measurable cost and paid back on every package produced at the line’s rated speed
- Compliance stopped depending on the sample, because the margin now covers the distribution tail rather than the mean
Scenario based on a Sailwin customer project; site-specific figures available on request during engineering review.
Frequently Asked Questions
Set Your Fill Target From Data, Not From Habit
Send your bottle drawing, container sample or target output. Our engineering team replies with a machine recommendation, mould assessment and factory-direct quotation within 24 hours.
Related Reading:
• Filling Machines and Monobloc Lines Overview
• Gravity vs Piston vs Flowmeter Filling
• 3-in-1 Monobloc Rinse, Fill and Cap Machines
• Isobaric Filling for Carbonated Soft Drinks
• Hot Fill Lines for Juice and Tea
• CIP Cleaning for Filling Lines
