What Is Loss-in-Weight Feeding?

Every continuous process has a ratio at its centre: how much of this, for every unit of that. When the ratio drifts, the consequences do not stay in the feeder — they show up as scrap, as rejected batches, as a product that fails a specification three steps downstream. Loss-in-weight feeding exists to hold that ratio, and to hold it without an operator watching.

The principle: weight loss as the setpoint

A loss-in-weight feeder is a weighing instrument with a screw underneath it. The hopper, the material inside it and the feeding mechanism are suspended on load cells, so the entire assembly is weighed continuously. The controller knows the total mass, samples it at short intervals, and calculates how many kilograms per hour are leaving the system. That calculated rate — not the screw speed — is the controlled variable. If the weight is falling faster than the setpoint, the screw slows down; if it falls more slowly, the screw speeds up.

Two things make this work in practice. The first is the feed factor: the controller learns how many grams each revolution of the screw delivers for this particular material, and it keeps updating that value as the material changes. The second is the refill cycle. When the hopper empties to a low-level threshold, it is refilled quickly, and during that refill the weight signal is no longer a clean measure of discharge. A well-designed feeder handles this by switching to volumetric control for the few seconds of refill, then returning to gravimetric control once the weight is stable again. How gracefully that transition happens separates a feeder that holds ±0.5% from one that claims it.

It is worth being precise about what an accuracy figure means. At PEGE we specify ±0.5% of setpoint on most platforms and ±1% on the smallest micro model, with digital weighing resolution down to 1/100,000 on the twin-screw range. Those numbers assume continuous operation of at least thirty seconds, a bulk density of 0.6 g/cm³, and a feeder calibrated against the actual material. A number without those conditions is not a specification.

Why weight, and not volume

The older alternative is volumetric feeding: run the screw at a fixed speed and assume that each revolution moves a known volume, which becomes a known mass once you multiply by bulk density. It is simple, inexpensive and perfectly adequate for materials that behave consistently.

The difficulty is that bulk density is not a constant. It changes between deliveries, it changes with moisture, it changes when a pneumatic conveying line aerates the powder on its way into the hopper, and it changes again after a few hours of vibration in a silo. A volumetric feeder cannot see any of this, so its actual output moves while its setpoint stays still. Loss-in-weight feeding measures mass directly, which means density drift is measured and corrected rather than assumed away.

How PEGE feeders differ from conventional weighing

The most common comparison is with batch weighing, or gain-in-weight batching, where material is added to a scale until a target weight is reached, then discharged. Batching is accurate and easy to audit, but it is intermittent by nature: it needs a buffer between the scale and the process, it delivers material in pulses rather than a continuous stream, and at low rates the cycle time becomes impractical. Loss-in-weight feeding delivers continuously, directly into the host machine, with no buffer vessel in between.

The second comparison is with conventional loss-in-weight feeders themselves, and here the differences are in the details rather than the principle. PEGE suspends the hopper on a cantilever beam with load cells matched to the working range instead of oversized for safety margin, so small weight changes are not buried in sensor noise. The drive is servo with planetary reduction, which removes the dead band a standard gearmotor introduces at low speeds. Control runs a speed loop together with a time-based correction, and the PI parameters re-tune themselves when a refill or a density change disturbs the signal. Powder platforms use a mirror-image U-shaped hopper that breaks arching before it starts, and the FLEXHOPPER moves cohesive material by massaging a flexible hopper wall from the outside, so nothing touches the product.

The practical consequence is that we can choose the feeding principle instead of defending the one we build. Six platforms and twenty-one models span 0.3 to 2,000 kg/h — single screw for granulate, twin screw for powder and fibre, micro twin screw for additives, flexible paddle for cohesive powder, liquid for additives and pre-melt products, vibratory tray for flakes and fragile material. When a customer‘s material suits a different principle from the one they enquired about, we say so.

Table 3   Loss-in-weight against volumetric and batch weighing

Where PEGE is taking loss-in-weight feeding

Three directions guide our development work, and they are directions rather than finished claims.

More precise. The limit on accuracy at low rates is usually the weighing system, not the screw. Higher-resolution digital load cells, better temperature and vibration compensation, and control that distinguishes real weight change from mechanical disturbance all push the usable range downward — which is what allows a feeder to hold a 0.5% additive stream at a few hundred grams per hour.

More intelligent. A feeder that knows its own feed factor can also predict when it will need refilling, flag a drift before it becomes a rejected batch, and report its condition without anyone visiting the line. Remote diagnostics over Ethernet are already standard on our units, with PROFIBUS-DP, Profinet and Modbus RTU available; the direction is to make that data useful to the customer‘s own system rather than merely visible.

More capable of meeting varied demand. Production is moving toward smaller batches, more formulations and more frequent changeovers, on the same line. That favours modular machines: a control platform, hopper and discharge element that can be changed on site, a feeder that can be re-specified for a new material without being replaced, and a supplier able to build the non-standard configuration rather than decline it.

What it comes down to

Loss-in-weight feeding is not complicated to describe: weigh what is leaving, compare it with what should be leaving, correct the difference. What is difficult is doing it for years, on a real material, in a real plant. That is where the engineering sits — and it is where we spend our time.