Weighing Systems Part II -Getting the Install Right

GG50 Weighing Series

Last post we covered what a scale and what has to be true for that number to mean anything. Gross, tare, net. Short ton, long ton, tonne. Truck scales, rail scales, hoppers, tower bulk weighers, and the draft survey you use when there's no scale to be had. All of that assumes the equipment itself was put in right. That's not always a safe assumption, no matter what kind of scale you're standing on.

This post applies across the board, not just truck scales. A pit-type truck scale, a hopper hanging from cells above a leg, a rail track scale, a tower bulk weigher cycling all day inside an export elevator, they're all precision instruments that got mounted onto structure, hooked to electronics, and left to do their job in real conditions. Every one of them can be wrecked by a bad install, and none of it shows up on a calibration certificate.

Since this reaches a global audience, it's worth saying plainly: not everybody's weighing grain on a digital scale to begin with. Plenty of the world still runs on a mechanical beam scale, and there's nothing wrong with that. A well-maintained beam scale with good reference weights is honest equipment. You'll also run into a hybrid version out there, a mechanical beam retrofitted with a single load cell mounted on the arm instead of a full electronic conversion. Cheaper than replacing the whole scale, and it can work fine, but it inherits every mechanical weak point of the beam itself, pivots, bearings, linkage wear, on top of whatever the one cell tells you. Whatever style of scale you're using, mechanical, hybrid, or fully electronic, the same rule applies. The reading is only as honest as the iron underneath it.

A quick note before diving in: I'm not a certified scale technician and haven't been formally trained in scale installation or service work. What follows comes from decades of operational experience around this equipment, conversations with the people who do that work for a living, and some research to fill in the gaps. Treat it as an operator's perspective, not an engineering reference.

One more thing before we get into it. When it comes time to actually install or move a scale, that's a job for a certified scale technician, not a crew with a tape measure and good intentions. The cost of doing it right the first time is nothing next to what a bad install costs you in bad data, or worse, in a scale that fails under load.

Get the Install Right Before You Trust the Number

Calibration gets the attention because it comes with paperwork. A certified test weight, a signed certificate, a sticker on the indicator. All that really tells you is the scale read right on the day somebody checked it. It doesn't tell you the scale was installed sound to begin with.

A scale that's installed poorly can still pass calibration. A cell that's binding, a deck that's flexing, a hopper that's not venting right, all of that can look fine on a calm day with an inspector standing there running a test weight through it. Load it up under real dynamic working conditions and it maybe off. Calibration only checks that the math is right on the day you run it. It doesn't check whether the install itself was done right. Get the installation solid first. Once that's right, the number on the screen takes care of itself.

Scale Increments and the Legal Error from Zero

Before going further, it's worth knowing that "accurate" has a legal meaning, and it's not zero error. It's error within an allowed margin.

Every legal-for-trade scale reads in fixed steps, called the scale division, and every legal-for-trade scale also carries an allowable margin of error built into the standard it's certified against. Neither one is the same across the board. A truck scale, a rail scale, and a bulk weigher can all carry different reading increments and different allowable tolerances, and the number can shift again depending on the scale's size, its classification, and which country's standard it's certified under. "Within tolerance" isn't one fixed figure you carry from scale to scale. It depends on what type of scale it is and where it's located.

We brought in multiple scale experts to go over our bulk weigh hoppers at the loading end. We checked the structural members' integrity and modified the venting between the hoppers. We checked for binding in the air hoses and conduit. We checked for worn cloth gasket material between sections that could be letting air or moisture in where it shouldn't, and again, possible binding. We ran static and dynamic testing on the whole system, and through all of this found very little wrong, and definitely not enough to explain the variances. We fixed what little we did find. Kept getting the complaints anyway.

After a couple of years, I made the trip to Europoort to see their side of it firsthand. Turned out they were still running old beam scales on a series of small hoppers with a single load cell arrangement and calibrating them with a five pound can of weights. That five‑pound can wasn’t wrong for checking the point at one spot; it was wrong for pretending that single low‑end check proved a full test across the full range. We'd spent real time and money chasing our own equipment down to the last gasket, and the mismatch had been sitting on the other end of the ocean the whole time.

That's the real lesson behind the increments and tolerance discussion above. Both ends of a trade can be technically "in tolerance" by their own local standard and still not agree with each other, because the two standards were never the same standard to begin with. Chasing your own mechanical problems first is always the right move. But if you've chased everything on your end and the numbers still won't reconcile, it's worth asking what the other end's equipment and testing standard actually look like before spending another dime on your own scale.

Load Cell Location & Mounting

A load cell only reads right if it's carrying the load straight down, evenly, with nothing else fighting it. Sounds obvious. It's also the first thing that goes wrong.

● Base mounted vs. hung. Most truck and rail scales sit on cells mounted underneath, taking the load straight down through compression. But you'll also run into hung, or suspension, mounted cells, especially on hoppers and tank scales, where the vessel hangs from the cells instead of resting on them. Same principle, opposite direction of pull, and the two get installed and troubleshot differently. Know which kind you've got before you go looking for a mounting problem, because a fix that makes sense on a base-mounted cell can be the wrong move on a hung one.

● Placement matters. Cells have to sit where the structure actually delivers the load to them evenly. Get the mounting points off even a little and one cell carries more than the others. The indicator doesn't know that. It just adds up whatever the cells report.

● Shimming has a right way and a wrong way. Shims fix a structure that isn't level or a cell sitting slightly high or low. Fine when done right. When somebody uses shims to force a stubborn reading into line instead of fixing the actual problem, that's how you bury an error instead of solving it.

● Over-tightening is its own failure. Check rods and mounting hardware get torqued to spec for a reason. Crank them down past that because it "feels more solid" and you restrict the tiny bit of movement the cell needs to flex. That restriction reads as resistance, and resistance reads as weight that isn't there.

One bad cell doesn't stay isolated either. Most scales sum several cells together, and one reading wrong drags the whole total off with it. Finding that one bad cell among four, six, or eight of them is no small job. Get the install right the first time and you can reduce the likelihood of lost time looking for a bad cell.

Binding — Working Fine Until It Isn'

Binding causes more unexplained weight complaints than anything else on this list, and it's the hardest one to catch, because a bound scale doesn't fail outright. It just lies steady, in a way that looks like a real number.

A load cell has to compress and rebound by a tiny, precise amount every time weight hits it. Anything that gets in the way of that movement, even a little, throws the reading off.

● Debris packs into the gap around a load cell or check rod, especially on a pit-type truck scale, and starts carrying part of the load itself.

● Rust and corrosion seize up check rods and hardware that are supposed to move freely.

● Misalignment, from a structure that's shifted or was never quite square, puts a side load on a cell that was built to take weight straight down.

● Bent check rods fight the vertical movement the cell needs.

● Side loading from a curb, a conveyor support, or ice pressing against the platform edge, adds force the cell reads as weight even though nothing extra is actually on it.

●  Snow and ice packed into the gap between the deck and the pit wall on an in-ground scale binds the same way debris does. Rubber flashing is commonly used to seal that gap and keep material from falling through underneath, but the rubber itself can freeze stiff and bind against the deck, which defeats the purpose it's there for. It's part of why an above-grade, elevated scale has a real edge: there's no pit gap to pack with ice and snow in the first place.

Binding doesn't spike or throw an error. It's a number that's off by a small steady amount, day after day, until the totals quit reconciling and nobody remembers when it started. Physically walk the pit and check the hardware on a schedule. Don't just trust the calibration sticker.

Next up: what happens after the install crew's gone home. Vibration, structural flex, weather, and the people with access to the scale, in "What Wears a Scale Down."

Grain Guy Fifty invites readers to share their thoughts, experiences, knowledge, and insights after exploring the post. Together, we can learn and grow as an industry!

Best,

Jim Voigt

"Grain Guy Fifty"