Detecting Double Fills: The Core Principle

A double fill is detected reliably only when you verify each dispense event, not just the finished cartridge weight. Total-weight inspection alone will catch gross overfills, but it cannot distinguish a genuine second dose from normal variation in the empty cartridge, crystal bulk density, or moisture pickup. The dependable approach combines three layers: a per-unit tare (empty cartridge weight), a per-dispense net weight from a load cell under or above the fill head, and a logic counter in the PLC that permits exactly one dispense per indexed position.

This matters most in polyphosphate cartridge production because the media is a brittle, friable crystal rather than a free-flowing granule. Fill weight scatters naturally, so the tolerance band you set for a checkweigher has to absorb that scatter — and the wider the band, the more easily a small double dose slips through.

What Counts as a Double Fill

The term covers several distinct faults, and they need different detection responses:

• Two full dispenses into one cartridge. The unit passes the fill station twice, or the auger fires twice for the same index. Net weight lands close to twice the target.
• Target dose plus a top-up. The cartridge receives its normal fill, then a second, smaller trickle-fill because the level looked low after vibration settling. Total weight may sit only 10–30% above target — the hardest case to catch.
• Manual addition. An operator tops up a cartridge by hand to reach a target weight, usually after a rejected unit is reworked.
• Phantom double counting. The unit is filled once, but a sensor or counter registers two events. Nothing is overfilled, yet the batch record and media reconciliation will not balance — and the system may then skip a cartridge that actually needs filling.

The last case is worth separating, because the fix is in the control logic rather than in weighing.

Primary Detection Methods

1. Two-stage weighing (empty tare, then gross weight)

Weigh every empty cartridge immediately before the fill station and every filled cartridge immediately after. The difference is the true net fill, and it removes the largest source of false alarms — variation in blow-molded or extruded cartridge shells. This is the single most effective upgrade for a line that currently checks only the finished weight.

Practical requirements: a stable weigh platform isolated from auger and vibratory vibration, a settling delay long enough for the load cell to stabilise, and a reject gate that removes the unit before capping. If the empty cartridge can hold residual crystals or moisture from a previous pass, the tare reading must be taken on a clean, dry shell.

2. Loss-in-weight load cells on the fill head

Mounting the hopper or the dispensing auger on load cells lets the controller measure how much media left the system per dispense. Each dispense event is then a discrete record with a timestamp and a net mass. A second dispense for the same index is obvious in the data even if the finished cartridge weight stays inside a wide band.

This method also captures short fills and blocked or bridged augers, so it usually pays for itself as a general fill-integrity check rather than a double-fill detector alone.

3. One-shot logic and per-index counters in the PLC

Software interlocks are cheap and catch the most common root cause: a repeat trigger. Useful elements include:

• A latched one-shot per index position, cleared only when the unit physically leaves the station (confirmed by an encoder position or a photo-eye).
• A comparison of dispense count against unit count at the end of each batch. Any mismatch is a flag, not a nuisance alarm.
• A reject-and-lock rule: once a cartridge is rejected, it cannot re-enter the fill station without an authorised reset.
• Removal of any manual “top-up” button from the operator panel, or at minimum an audit log entry each time it is used.

4. Mass balance across the batch

Record the mass of media consumed and the number of cartridges produced. If the average fill per cartridge drifts upward over a shift, double fills or manual top-ups are occurring somewhere, even if individual reject rates look normal. This is a slow but very sensitive indicator, and it needs no new hardware.

Secondary and Backup Detection

Fill-height or freeboard measurement

A laser displacement sensor, ultrasonic sensor, or vision camera measuring the distance from the cartridge rim to the crystal surface detects overfill independently of weight. Accuracy is limited by the irregular surface of polyphosphate crystals, dust in the air, and translucent or tinted cartridge walls, so it is best used as a confirmation signal rather than the primary arbiter. Its main value is catching the case where a double fill coincides with a low-density crystal batch, keeping total weight near the expected value while the bed height is clearly wrong.

Dispense signature monitoring

Auger motor current, dispense duration, and vibratory feeder amplitude all change when the auger discharges into an already-full tube. Reading these traces costs nothing extra if the drive already reports them, and a repeat dispense into a full cartridge typically shows a distinct shift. Treat this as an early-warning trend, not a pass/fail gate.

Go/no-go gap gauge before capping

A simple mechanical gauge that checks freeboard height at the capping station is a low-cost catch-all. It will not tell you why a unit is overfilled, but it stops overfilled cartridges from being capped and shipped.

Sampling with destructive verification

Periodically weigh a filled cartridge, empty it completely, and weigh the empty shell. This confirms that your inline tare assumption is still valid and that crystals are not clinging to the tube wall. It is also the reference method for calibrating everything else on the line.

Comparing the Options

Method	Detects	Strength	Limitation
Empty tare + gross weigh	True net fill per cartridge	Removes shell-weight variation	Needs two stable weigh points
Loss-in-weight at fill head	Each individual dispense	Catches repeat dispenses directly	Vibration and cable stiffness affect accuracy
PLC one-shot logic	Repeat triggers and phantom counting	Very low cost, fast to implement	Cannot detect manual top-ups
Mass balance per batch	System-level overfill trend	No hardware required	Delayed, not unit-level
Height or vision check	Physical overfill of the bed	Independent of weight	Surface irregularity, dust, wall colour
Gap gauge before capping	Gross overfill	Simple and robust	No data, no root cause

Before You Blame a Double Fill

An over-target cartridge weight is evidence, not proof. Several other factors push net fill weight upward, and confusing them with a double fill leads to unnecessary line stoppages:

• Crystal bulk density variation. Different production lots of phosphate crystal can differ in density and in the proportion of fines. Fine-rich material packs more mass into the same volume.
• Moisture uptake. Polyphosphate crystals are hygroscopic. Media left in an open hopper over a humid weekend can pick up water, adding mass that is not extra crystal.
• Compaction after vibration. Settling tables and vibration stations change bed volume without changing mass; if the filler compensates by volume, mass will drift.
• Static charge. Fine particles clinging to the tube wall can be weighed but not properly bedded.
• Stale tare values. If the system assumes one average empty-shell weight, a change in shell supplier or mould will shift every net reading.

Confirm the cause before adjusting the filler. The destructive sampling check and a comparison of dispense-event records to unit records will usually settle the question within one shift.

A Practical Setup Checklist

1. Define the target net fill weight and the acceptable range from the product specification — not from what the line currently achieves.
2. Measure the empty-shell weight distribution over at least 30 units. If the spread is large relative to your tolerance, individual tare is mandatory.
3. Confirm the crystal’s bulk density and particle size distribution with your media supplier, and re-check on each incoming lot.
4. Install the one-shot logic and dispense counter first; it costs little and removes the most common root cause.
5. Add loss-in-weight or two-stage weighing, and log every event with a unit identifier for traceability.
6. Set the reject gate to act before capping, and lock rejected units out of the fill station.
7. Chart net fill weights per shift. A bimodal distribution — two separate peaks — is the clearest statistical fingerprint of double filling.
8. Reconcile media consumption against units produced at the end of every batch.

Note that the media itself influences how tight the detection can be. Polyphosphate crystals marketed under names such as Siliphos are not all identical: particle size, density, friability, and coating differ between suppliers and grades. Ask for the actual grade data rather than assuming two phosphate crystals will fill the same way.

Frequently Asked Questions

Can a checkweigher alone detect double fills?

Only large ones. A checkweigher measures total weight, so its tolerance must accommodate empty-shell variation and crystal density variation. Any double dose small enough to fall inside that band passes through. Two-stage weighing or per-dispense load cells close that gap.

How do I tell a double fill from a dense crystal lot?

Compare dispense-event records with unit records and check bed height. A double fill shows two dispense events for one unit, or a bed height clearly above the target freeboard. A dense lot shows a single dispense event with a normal bed height and a heavier net weight than the previous lot.

Is overfilling actually a functional problem?

Yes. Reducing freeboard changes how water contacts and passes through the bed, which can raise pressure drop across the cartridge. Compaction and fines migration can also restrict flow. Overfilled cartridges may also stress end caps and housings during handling. The exact effect depends on cartridge design and operating conditions, and should be confirmed by flow testing rather than assumed.

Do polyphosphate cartridges soften water?

No. Polyphosphate media are used for scale inhibition — they interfere with the crystallisation of hardness salts. That is a different function from ion-exchange water softening, from filtration, from scale removal, and from corrosion inhibition or disinfection. Cartridge performance claims should always be stated against the specific function the media is intended to perform.

Where should the detection point sit on the line?

As early as possible, and always before capping. A defect caught at the fill head can be corrected or the unit rejected cleanly; the same defect found after capping usually means scrapping the cartridge and the media inside it.

Specifying Media for a New or Reworked Line

Detection quality is limited by how consistent the incoming crystal is. If you are selecting or re-specifying polyphosphate crystal for a filling line, the information that actually determines fill repeatability is bulk density and its lot-to-lot range, particle size distribution, fines content, friability, and moisture behaviour in storage. ENVO CHEMICAL can review these parameters against your cartridge format and target fill weight — share your cartridge dimensions, target net fill, line speed, current filling equipment, and the water-treatment objective the cartridge must meet, and we will confirm which specification points need to be verified for your application.