Weighing Systems Part III -What Wears a Scale Down

GG50 Weighing Series

Last post covered getting a scale installed right in the first place: what "accurate" legally means, mounting the load cells correctly, and binding, the issue that can hold a scale off true without ever throwing an obvious fault. A scale that clears all of that is sound on day one. Staying sound becomes the challenge, and it's the one that runs for the rest of that scale's working life.

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, service,  move a scale, that's a job for a certified scale technician. 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.

Everything below is what goes to work on a correctly installed scale after the install crew's gone home: the air moving through it, the vibration around it, the structure under it, the weather on top of it, and the people with access to it.

Air Relief

Hoppers and bulk weighers fill and dump in cycles, and every cycle moves a slug of air along with the grain. A hopper that isn't vented right builds pressure as it fills and pulls a vacuum as it dumps, and that pressure pushes on the vessel walls the same as weight would. The cells can't tell the difference.

You'll see it as cycle weights that don't repeat, cycles that run slower than they should while the system fights its own pressure, and a total that changes depending on how fast you're cycling that day. Same grain, different cycle speed, different number. That's a vent problem before it's ever a calibration problem.

How many stages the system has changes how much air you've got to manage. A two-stage system, upper hopper feeding the scale hopper, which discharges straight to conveyance, only has to equalize air between two vessels: the upper hopper and the scale. A three-stage system, upper hopper, scale hopper, and a lower hopper below the scale before the grain hits conveyance, adds a third vessel to the air path, and now you're equalizing pressure across all three, not two. More vessels in the chain means more places for a venting mismatch to hide, and more work to track down which junction is actually the problem when a cycle weight comes in wrong.

Bad venting doesn't just add phantom weight, either. It slows the scale down. A hopper still fighting pressure or vacuum takes longer to settle to a stable reading, which stretches out the settle time the cycle needs before it can weigh clean. On a slower-cycling elevator loadout that's an annoyance. On a system running at unit train pace, where settle time is of priority, a venting problem eats into time you didn't have to spare.

Vibration

Load cells are built to catch tiny changes in force. That's the whole point of them. It also means they pick up vibration from a leg running, a conveyor under load, a fan, a compressor, even truck traffic on the scale next door.

Most systems filter that out fine on a static reading. On a fast-cycling process scale or a belt scale, it doesn't always average out clean. You'll see it as jitter, a reading that won't quite settle, or small differences between cycles that should read identical.

Here's the part people miss: vibration isn't a one-time install check. You can get the environment dialed in fine when a scale goes in, no problem for months, and then something around it changes. A fan on the dust collection system next to your weigh hopper goes out of balance and starts putting a new frequency into everything it's bolted to. A bearing on a nearby leg starts going bad and the vibration it throws off climbs a little more every week before anybody notices. You add a second truck platform scale right alongside the first one to speed up traffic, and now every truck rolling across scale two is putting vibration into scale one that was never in the picture when scale one was installed and calibrated. None of that shows up as a wiring problem or a mechanical failure on the scale itself. It shows up as a reading that's gotten a little less steady than it used to be, and the cause is sitting somewhere else entirely. Isolating a scale from vibration sources isn't something you do once at installation and forget. It's something you keep checking, because the equipment around the scale keeps changing, wearing, and getting added to, even when the scale itself hasn't been touched.

Structural Flexing

A scale doesn't work on its own. It works because it's sitting on a foundation or a frame rigid enough to deliver the load to the cells without absorbing any of it itself. If that structure flexes under load, the cells are chasing a moving target.

This shows up most on bulk weighers and hoppers set on steel that was sized for the equipment's dead weight, not for the dynamic load of a full hopper cycling all day, every day, for years. A deck that flexes even a little changes how load transfers to the cells depending on how full it is. That's an error a bench test won't catch. It only shows up under real production, which is exactly the condition nobody's checking once they've decided the scale's fine because it calibrated clean.

This is especially critical on a suspended hopper scale. A base-mounted scale only has to answer for vertical load, straight down into the cells. A hung hopper is different. The support steel isn't just carrying weight, it's also fighting the horizontal sway and side-load a hanging vessel puts into its structure as it fills, settles, and dumps. Both the vertical support and the horizontal bracing have to be engineered for that, not just the vertical member sized for dead weight. Undersize or overlook the horizontal support and you get exactly the same problem as an undersized vertical member: the structure absorbs load and flexes instead of delivering it cleanly to the cells, except now it's happening in a direction a lot of people never think to check.

Unit Train Loadout Towers — A Special Case

Tower bulk weighers deserve their own callout, because a unit train loadout tower stacks nearly every problem in this post on top of each other at once, running flat out, for hours at a time.

A unit train isn't loaded car by car with time to think about it. It's 100, 110, sometimes 130-plus cars moving through a loadout at a pace set by how fast the railroad wants that train off the property. That means the tower's cycling about as fast as a bulk weigher can cycle: fill, weigh, dump, repeat, over and over, with a settle time that is  never fast enough. A settle time that's fine on a slower-cycling elevator loadout can be too short here. The weighment happens before the reading's fully steadied out, and you get a cycle weight that's technically a number but not the true one.

Rail vibration compounds it. The equipment moving and holding the cars in place through the loadout, whether locomotive(s) or car indexer, puts its own vibration into the tower, on top of whatever the tower's own leg and conveyor equipment is already generating. Add in the general jostle of a string of loaded cars advancing through the loadout, and the tower's absorbing more vibration input than almost anywhere else in the operation. It's the vibration problem from earlier in this post, except now it's happening at the fastest cycle rate the equipment runs and with the least settle time to filter any of it out.

The structural demands go up right along with it. A tower cycling at unit train pace, day after day during loading season, is putting more fatigue cycles through its mounting steel in a week than a slower-cycling facility sees in a month. Structural flexing that would take years to show up somewhere else can show up in a season here. And every one of those fast cycles is also moving air fast, so a tower that's marginal on venting will show it here first, with cycle weights that swing depending on how hard the loadout's running that hour.

None of that means unit train towers can't weigh accurately. It means the margin for error in installation, mounting, venting, and vibration isolation is smaller than anywhere else in the operation, because the equipment's being asked to do its job faster and with less room to settle than anywhere else in the chain. If there's one piece of weighing equipment in a facility worth an extra hard look at install time and an extra frequent physical inspection afterward, it's the tower feeding the unit train.

Large export facilities have similar issues with their bulk weighing systems as train loaders, but given the loading rates required for vessels you may have multiple scales working side by side.

Environmental Intrusion

Scales live outside, or close enough to it, and weather doesn't care about your calibration schedule.

● Water gets into pits, junction boxes, and cell housings, and it changes the electrical characteristics of a cell's wiring before you ever see visible corrosion.

● Ice and snow buildup in the scale gap is a bigger problem than people give it credit for, and it's covered in more detail earlier in this series as a binding mechanism. In practice, it means any operation running scales through a real winter needs somebody keeping that gap clean on a schedule, not just plowing the approach around it.

● Mud and dust pile up in pits and around hardware, adding weight that isn't grain and feeding the same binding problems covered above.

● Rodents get into junction boxes and chew cable insulation, which shows up as intermittent, maddening electrical faults that look like anything except a wildlife problem, until somebody finally opens the box.

● Seasonal drift is the slow version of all this. Temperature affects cell output a little, ground heaves and settles with freeze-thaw, and a scale that read true in July can be off a real amount by January, for reasons that have nothing to do with the electronics.

None of this is exotic. It's weather doing what weather does to anything left outside for years. A scale that doesn't get walked and inspected regularly will pick up every one of these problems at the same time, quietly, until they show up together as a total nobody can explain.

Grounding & Electrical Protection

A load cell puts out a very small electrical signal, and that signal has to travel through cable, junction boxes, and an indicator without picking up noise along the way. Bad grounding lets noise into that path. A real electrical event, lightning being the extreme case, can just kill it outright.

I saw this firsthand at a barge loading facility on the Illinois River, back in the late 1980s. This was before updated surge protection was standard equipment. The truck scale on that site had a metal deck, and over one year, lightning found it five separate times. Five strikes, five outages, five trips out to figure out what got fried and get the scale running again. The fix was pouring a new deck in concrete. Once that was done, the outages stopped.

For a while I read that as the metal deck drawing the strikes in, same as the old myth about lightning rods pulling lightning toward a building. Neither one is true. Height, shape, and isolation decide where a strike goes, not the material sitting there. A rod on a barn or an elevator leg doesn't attract a strike, it just gives one a clean path down once it's already coming. What almost certainly changed at that scale wasn't how often the spot got hit. It's that a steel deck sitting right over the load cells and wiring gave any strike nearby a straight shot into the electronics. Concrete broke that path. Add the improved grounding and surge protection that came with the rebuild, and you've closed off both the direct and indirect ways lightning gets into the electronics. Fewer failures showed up. I mistook that for fewer strikes.

Scale Security

Everything above is about a scale mis-reading to you by accident. It's worth talking about the kind that lies on purpose, because that's not new either, and it hasn't gone away, it's just changed shape.

On old mechanical beam scales, tampering was a physical act, and inspectors have found it done in some remarkably direct ways. A guy sitting in a chair underneath the scale deck, out of sight, riding a sliding weight or leaning on a linkage to shift the balance point at exactly the right moment. Crude, but it worked, because the whole system relied on the beam and the weights being left alone between calibrations, and nobody was watching what happened underneath. That's the same trust problem this whole series keeps coming back to. The scale's only honest if nobody's got their hand on it.

Digital scales moved that risk somewhere else instead of removing it. You don't need a guy in a chair anymore. You need access to the indicator's configuration settings, the load cell junction box, or, on a networked system, the software and the connection carrying that weight data off to inventory or billing. Span adjustments made without authorization, a junction box opened to swap a resistor or tamper with the summing card, a networked indicator accessed remotely and its calibration parameters nudged, all of that produces a number that looks completely normal on the ticket and is wrong on purpose. It's harder to spot than the old chair-under-the-scale trick, not easier, because there's no guy to catch sitting there.

The controls that matter haven't changed as much as the technology has. Physical security on the scale house, the junction boxes, and anything that gives access to calibration settings. Password protection and access logging on any indicator or system that's networked. Sealed and tamper-evident calibration adjustments, so a change leaves a mark even if nobody's standing there to see it happen. And the same cross-checking that catches an honest mechanical error, comparing one scale's numbers against another, catches a dishonest one too. A rigged scale that's consistently off will show up in a cross-check the same as a bound load cell will. The tampering method changed. The way you catch it mostly hasn't.

Operator-Real Takeaway

Most weighing problems that get blamed on calibration aren't calibration problems. They're installation, maintenance, and environmental problems wearing a calibration-shaped disguise. A scale can pass every certified test on the books and still lie to you under real conditions if what's underneath it isn't sound.

Fix the mechanical side first. Mount the cells right, whether they're base-mounted or hung. Keep binding out of the system. Let the hoppers breathe. Isolate the vibration and keep checking, since the equipment around your scale keeps changing even when the scale doesn't. Make sure the structure isn't flexing. Keep the water, ice, and rodents out. Ground the electronics. Keep an eye on who has access to calibration settings, physical or digital. Do that and the number on the indicator takes care of itself. Skip it and no amount of calibration paperwork saves you.

Next up: how you keep a scale honest over time. Calibration schedules that actually mean something, cross-checking one scale against another, and legal-for-trade certification requirements, which vary more by region than most people realize.

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"