Why Point Level Sensors Fail on Open Stockpiles

Reliable stockpile volume measurement is one of the harder problems in bulk material handling, and the instrument you choose at the design stage largely decides whether the numbers you report later are trustworthy. On open outdoor heaps, a single point level sensor measures the height of the material at one fixed location. That works well inside a tank or a regular silo, but an open pile is a different physical object: its surface is irregular, its footprint is large, and its shape shifts every time material is added or drawn down. This article compares the two approaches on technical grounds so you can match the sensing method to the storage geometry rather than forcing one device to do a job it was never designed for.

The conclusion is not that point level sensing is obsolete. It remains the right answer for many enclosed vessels. The point is narrower: a method that samples one height cannot describe the volume of a surface that varies across its whole area. Understanding why is the key to selecting correctly.

What a point level sensor actually measures

A point level instrument — a radar, guided wave, or ultrasonic device — reports the distance from a fixed reference at the top of a vessel down to the material surface directly beneath it. Subtract that distance from the known vessel height and you get the fill level along one vertical line. In a cylindrical silo or a rectangular bin with vertical walls, that single height is enough: the cross-sectional area is constant and known, so volume is simply area multiplied by the measured height. The geometry does the heavy lifting, and the sensor only has to nail one number.

This is exactly the application a silo radar level transmitter is built for. Inside a closed vessel with a predictable cross-section, a well-installed silo radar level transmitter delivers a clean, repeatable reading with very little signal processing. The walls constrain the material, the headspace is controlled, and the surface — while it may cone or rathole — stays within a small, well-understood envelope. There is nothing wrong with the device. The assumptions it relies on simply hold true in that setting.

Why those assumptions break on an open pile

Move the same sensor over an outdoor stockpile and every assumption it depends on collapses at once.

First, there is no fixed cross-section. A pile of coal, ore, aggregate, woodchip, or grain sits at its natural angle of repose and spreads into a cone, a windrow, or an irregular mound with multiple peaks and hollows. The “area” you would multiply by a height is unknown and changes constantly, so the core point-level equation no longer has the inputs it needs.

Second, one height is not representative. A point sensor reads the material directly under it. On a heap, the spot beneath the sensor might be a peak, a valley, or the slope between them. The same pile can return wildly different readings depending on where the beam happens to land, and none of those single readings tells you anything dependable about the rest of the surface.

Third, the shape itself is the variable you care about. Loaders carve into one face while conveyors build up another, so the pile is asymmetric and in constant motion. Coverage from a single fixed point — or even a handful of points — leaves most of the surface unobserved. You are sampling a few lines through a body whose entire three-dimensional form is what determines its volume. The information needed for stockpile volume measurement is spread across the whole surface, and a point method, by definition, cannot capture it.

The result is not a small error you can calibrate out. It is a structural mismatch between what the sensor reports and what the application requires.

How 3D LiDAR area scanning closes the gap

Area scanning attacks the problem from the opposite direction: instead of sampling one height, it measures the whole surface. A 3D LiDAR stockpile measurement system sweeps a laser across the pile and records the return from thousands of points, building a dense point cloud of the surface. From that point cloud the system reconstructs the heap as a three-dimensional model and computes volume directly against a surveyed base plane or floor reference — no constant cross-section is assumed, because none is needed.

Because the method captures the surface rather than a single line, it adapts naturally to irregular and shifting shapes. Cones, multiple peaks, drawn-down faces, and asymmetric mounds are all just geometry the point cloud already describes. As material moves, the next scan re-measures the new surface and recomputes the volume from scratch, so the reading tracks the pile instead of fighting it.

Accuracy depends on calibration, scanner placement, surface conditions, and the specific model, but properly set up systems typically achieve volume accuracy on the order of 1–3 percent after calibration. Scanning range is model-dependent and spans roughly 0.5 to 100 metres across the product family, which covers everything from a compact indoor bay to a large outdoor yard. These figures are representative rather than guaranteed; the right number for a given site comes from matching the model and mounting to the pile dimensions and material. Treated honestly, area scanning gives you a measurement whose method matches the object being measured — which is the whole reason it succeeds where a point sample cannot.

When to use point level versus 3D LiDAR

The selection rule follows directly from the geometry.

Choose a point level sensor when the material is held in a regular, enclosed vessel with a known and constant cross-section: a silo, a tank, a vertical-walled bin, or a hopper. There the single-height method is accurate, low-cost, simple to install, and entirely fit for purpose. Reaching for a scanning system in that case adds complexity with no measurable benefit.

Choose 3D LiDAR area scanning when the material sits in an open or large stockpile, an irregular heap, a long bunker, or any storage where the surface shape is not fixed and the footprint is large. There the surface is the measurement, and only a method that captures the whole surface can return a defensible volume.

Many real sites run both. Enclosed feed silos stay on point level radar; the open yard or the long shed runs on LiDAR. The two technologies are complementary tools for different geometries, not competitors for the same job. The mistake to avoid is using the enclosed-vessel method on an open pile and then trying to explain away the error — the limitation is in the method, not in the calibration.

Getting the selection right for your site

If you are scoping a new installation or troubleshooting volume readings that never seem to settle, start from the storage geometry and work back to the instrument. A constant, enclosed cross-section points to point level. A variable, open, or large surface points to area scanning. Material properties, mounting access, dust, and reporting requirements then refine the model choice within whichever method the geometry selects.

When the application is an open or irregular stockpile and you need defensible volume figures for inventory, reconciliation, or process control, our team can review your yard dimensions, material, and layout and recommend a suitable LiDAR configuration. Request a stockpile measurement review and we will work through the technical fit with you, including expected accuracy for your specific geometry.

—

Related reading: for a broader overview of selecting level and volume instrumentation across enclosed and open storage, see our technical insights library.