Weighing Systems Part IV - Proving the Number Right

GG50 Weighing Series

Part II covered getting a scale installed right. Part III covered what goes to work on a good install after the crew goes home. Air. Vibration. Structure. Weather. And the people with access to it. Together that gets you a scale that is mechanically sound and holding up under real conditions. That is the foundation for accurate weighing. It is not the same as proof.

Proof is what this post is about. Calibration. Testing. Cross checking. And certification. These four things turn “this scale is probably fine” into a number you can put your name on. This closes out the series. It is the part that runs for the rest of that scale’s working life.

A quick note before diving in. I am not a certified scale technician. I have not 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.

Calibration Discipline

Calibration is a schedule. Not an event. Too many operations treat it like a box to check once a year when the inspector is due. Then they forget about the scale until the next visit. A scale that is mechanically sound on day one can still drift. A scheduled check catches that before it costs anybody money.

What that schedule looks like depends on the equipment.

• Truck and rail scales get checked with certified test weights. Or a certified test car on rail. Run the weights across the full range of the scale. On large bulk weighers this is typically done as a build-up (or substitution) test when the available certified weights are only a fraction of capacity. Not just one point near the low end. One reading at one weight only tells you the scale is right at that weight. It tells you nothing about how it behaves loaded to capacity.

• Bulk weighers and tower systems get verified against cycle weight. That is the number the scale actually runs with. A static test weight tells you the scale reads right sitting still. It does not tell you the cycle is repeating clean at production speed. That production number is what everybody downstream is trusting.

• Belt scales get both a static test and a dynamic material test. Sometimes called a chain test. Or a test with a known weight run across the belt at operating speed. A belt scale that is dead accurate sitting still can still read wrong once material is actually moving across it.

None of that replaces a walk around. Do a physical inspection on a set schedule. Check for binding. Debris. Buildup. Worn cabling. And the other things covered in Parts II and III. The inspection catches problems a calibration check alone will miss. A scale can still zero and span correctly with a mechanical problem quietly building underneath it. Calibration and inspection are not the same task.

Anything that changes the scale’s environment or mechanical condition earns an extra check outside the normal schedule. After a repair. After structure gets modified nearby. After a hard freeze thaw cycle. After somebody mentions the numbers have felt off. Waiting for the next scheduled date when you already have a reason to suspect a problem is how a small drift turns into a season’s worth of bad tickets.

Testing Methods

A calibration check is not one test. It is a handful of them. Each one is looking for something different.

• Zero check. Does the scale return to a clean true zero every time it is unloaded. Within the tolerance its own standard allows. This is the baseline every other reading gets measured against. Binding, buildup, and debris can hold a scale off zero without ever throwing an obvious fault.

• Span check. Does the scale read correctly across its working range. Not just near zero. A certified test weight or test car run at multiple points—increasing and then decreasing—catches a scale that is accurate light and wrong heavy, or the reverse. It also reveals hysteresis (the difference in reading for the same weight depending on whether you are loading or unloading).

• Cycle weight consistency. For bulk weighers and tower systems. Does the same fill produce the same reading cycle after cycle. At the pace the equipment actually runs. This is where a venting problem or a settle time problem shows up first.

• Flow tests. For belt scales. Does the totalized number match a known quantity run across the belt at real operating speed. Not a static bench number.

Run all of them on schedule. Not just when something feels wrong. A scale that passes a zero check but has not had its span checked in a year is only half tested. And half tested is functionally untested for the half you skipped.

One more thing worth testing directly is settling time. Or cycle time. Every scale needs a certain amount of time to stop moving and give you a true reading. That number can drift without anybody touching a setting. Watch it on its own. Not just the weight it produces. A creeping settle time usually means something is fighting the cell before it rests. Binding. A venting problem. Or structure letting the load bounce instead of settle. Shortening it to push more cycles through a shift trades accuracy for speed. A weight taken before the reading has steadied out is a number. Not the true one. Track settle time the way you would track zero or span.

Cross Checking Scales

Calibration tells you a single scale is honest against a known reference. Cross checking tells you your scales agree with each other. It is the cheapest insurance in grain handling. It costs nothing but attention. And it catches problems calibration alone will miss.

The easiest cross check to run is the monthly inventory reconciliation. Your own inbound weight against your own outbound weight. Everything that comes onto your books eventually leaves it. Your inventory system should already be identifying handling and moisture shrink. If you see a trend where that shrink starts expanding, start with your scales. The gap between the two should track close and consistent. Not drift wider every month. Start there before you get fancy anywhere else.

A few practical pairings:

• Inbound truck weight against the tower or bulk weigher totals for the same period. They will not match to the pound. But they should track closely and consistently. Not drift apart a little more every week.

• Outbound truck weight against rail loadout totals when the same grain is moving both directions through a facility. A gap that is stable is probably just handling loss. A gap that is growing is a scale telling you something.

• Belt totals against bulk weigher totals on the same run. These are two different technologies measuring the same grain. When they stop agreeing, that disagreement is usually cheaper to chase down early than late.

If you have more than one location, use it. Or a neighbor a few miles down the road who will return the favor. Run a truck across your scale and then across theirs. You have an independent check that has nothing to do with your own equipment, environment, or calibration schedule. Multi location operations should be doing this between their own sites as a matter of course. If you are a single site, a friendly competitor or neighboring elevator willing to cross check a load once in a while is worth more than another calibration certificate. It tests something a certificate cannot. Whether your number and the outside world’s number actually agree.

And do not stop at the inbound and outbound scales. The bench and platform scales in the grading room need the same discipline. The ones used for moisture, test weight, and grade factors. It is easy to treat a small lab scale like an office appliance. But a grade built on a scale that has drifted is just as wrong as a load out ticket built on a bad tower weigher. And harder to notice. Nobody is cross checking a grading scale by habit. Put it on the same calibration schedule. The same zero and span checks. And the same walk around as everything else in this series. It does not move a semi load of grain. But it decides what that grain is worth.

Legal for Trade Certification by Region

Post 1 introduced the idea that certified legal for trade scales are the modern stand in for the handshake Joseph and the Byzantine traders never had. Here is the fuller picture on who stands behind that certification. Region by region. This is worth a quick check before you take it as gospel. Agencies update standards. This kind of regulatory detail can shift.

• United States. Domestically, state weights and measures offices inspect and certify commercial scales. They work under standards laid out in NIST Handbook 44. That handbook is revised and reissued on a regular cycle. Export weighing, especially at the terminal level, brings in FGIS. That is the Federal Grain Inspection Service under USDA. Covered in more detail back in Post 1.

• Canada. Measurement Canada is the federal authority. They approve and inspect trade scales under the Weights and Measures Act.

• European Union. Member states enforce their own inspection regimes. Device design and type approval trace back to OIML. That is the International Organization of Legal Metrology. Their recommendations underpin the EU’s Measuring Instruments Directive.

• Elsewhere. Most countries run their weighing certification through a national metrology institute. Often built around the same OIML framework the EU relies on. The specifics of inspection frequency, tolerance, and enforcement vary by country. Enforcement and inspection intervals can be significantly looser in practice in some markets. That makes cross checks and reconciliation even more important when you are trading across borders.

The common thread is simple. None of these agencies are in the business of trusting equipment. They are in the business of making equipment trustworthy. On a schedule. With a paper trail. So two parties who have never met can still do business off a number neither one generated. That is Joseph’s problem. Solved with an inspector and a certificate instead of a shared reputation.

What Certified Accurate Really Means

Worth being clear on what certified accurate actually means. It trips people up. A certified scale is not a scale with zero error. It is a scale operating within a set of allowed tolerances. Zero tolerance. Span tolerance. Repeatability tolerance. Each one is checked and permitted on its own.

A scale can sit at its allowed limit on zero. Its allowed limit on span. And its allowed limit on repeatability. All at the same time. And still pass certification on paper. Each deviation is legal by itself. Nothing requires those deviations to cancel each other out. If they happen to stack in the same direction instead, a scale that is fully certified on every line of the test can still read meaningfully off true on a real load. There is no line on the certificate that catches it. The test checks each tolerance in isolation. Not what happens when several of them lean the same way at once. That is one more argument for cross checking. Certification tells you the scale passed its tests. It does not tell you those legal permitted margins did not all lean the same direction today.

It helps to think about the effect in real terms. On a big commercial scale, a set of small legal biases can add up to several bushels or more on a single truck. And many more over a season. The paperwork still says in tolerance. The settlement still depends on the number.

When the Gap Is With the Other End

Everything above assumes the mismatch is in your own operation. Sometimes it is not. If a shipper or receiver’s numbers show a consistent variation against yours, that is a different problem. Not a one off short load. A gap that shows up trip after trip. Same direction.

This is the Europoort story from Part II. We chased our own bulk weigh hoppers for the better part of two years on a running complaint of light loading. We inspected structural members. Venting. Binding in the air hoses. Gasket material. Static and dynamic testing across the whole system. Found almost nothing wrong. The complaints kept coming anyway. It turned out the receiving end was still running old beam scales. Calibrated with a five pound can of weights. Never adequate for proving the full range of what those hoppers actually weighed. Two operations. Both technically in tolerance by their own local standard. Neither standard was ever really the same standard to begin with.

So work your own side first. Every time. Rule out binding. Mounting. Vibration. Structural flex. Drift. Everything covered in Parts II and III. Before pointing a finger anywhere else. But once you have genuinely exhausted your own equipment and the gap is still there, ask what is actually on the other end. What scale. What class. Certified against what standard. Tested across what range. A five pound can of weights proves one point. Not a hopper full of grain. That is an uncomfortable conversation to start with a repeat customer. But it is cheaper than quietly eating a variance for years while assuming the fault is yours.

Troubleshooting Drift

Every one of the previous three posts eventually points back to the same question. Why did the number change. And where do you start looking. A working approach, in order:

• Is it one scale or several. One scale points to something local. Mounting. Binding. A bad cell. The mechanics covered in Part II. Several scales drifting together points somewhere shared. A power supply. A grounding issue. Or a facility wide electrical problem.

• Is it one direction or all readings. A scale that is only off heavy, or only off light, behaves differently than one that is noisy in both directions. The first often points to binding or a structural load path issue. The second points more toward vibration or electrical noise. Covered in Part III.

• Is it tied to a season, a weather event, or a time of day. Seasonal drift and freeze thaw movement showed up in Part III. If the number moves with the thermometer or with a hard freeze, you are looking at environmental intrusion or structural flex before you are looking at the electronics.

• Has anything changed nearby. A new piece of equipment installed next to the scale. A fan that has gone out of balance. A repair made somewhere upstream. Vibration and structural loading do not stay fixed once a scale is installed. The cause of a new problem is often sitting somewhere other than the scale itself.

Work that list before assuming the scale itself has failed. Most of what looks like a calibration problem traces back to something mechanical or environmental wearing a calibration shaped disguise. That was the whole point of Part III. Calibration and cross checking tell you something is wrong. Parts II and III are usually where you find out what.

Operator Real Takeaway

Calibration keeps a scale honest against a known reference. Cross checking keeps your operation honest against itself. Certification keeps regulators, buyers, and the party on the other end of a contract satisfied the number was not just made up. None of the three does the other’s job. Skipping any one leaves a gap the other two will not cover.

Run the schedule. Test the full range. Not just one point. Compare your scales against each other. Not just against a certificate. And remember what this series has circled since Joseph stood over a storehouse he could not afford to miscount. Every number on a scale ticket is a claim somebody is making about reality. Your job is to make sure that claim holds up.

Closing Out the Series

That is weighing. Start to finish. From a Byzantine trader trusting a jar count. To a load cell converting force into an electrical signal. To a scale that has to prove itself over and over before anybody downstream can trust the number it hands them. Four posts. And the through line never really changed. Trustworthy weights are not a nice to have. They are the foundation everything else in this business gets built on.

This series picks up right where the sampling series left off. Sampling told you what you have actually got based off a representative sample. Moisture. Foreign material. Damage. The quality of the load, before you can trust anything built on top of it. Weighing told you how much of it you have got. Between the two, that is the pair of numbers nearly every settlement, contract, and audit in this business comes down to. Next up is inspection and grading. The piece that decides what all of it is actually worth.

One last reminder before you go. Do not let security slide off the list just because this post ended on calibration and certification. Part III covered how a scale can be made to lie on purpose. A rigged reading can pass every zero and span check on the books. Physical security on the scale house and junction boxes. Password protection and access logging on anything networked. And tamper evident calibration seals. These belong on the same routine as everything else in this series. Not a separate conversation you have only after something has already gone wrong. A clean trail of who touched what, and when, is part of defending your tickets in a dispute. Just as much as the calibration certificate itself.

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