An out-of-tolerance calibration result is not just a failed sticker.

It can affect product quality, safety decisions, customer acceptance, compliance records, past measurements and future work.

The original JM Test article explains the risk clearly: when an instrument does not meet specification, the owner may need to take action, possibly including review or recall of work already performed, and the manufacturer or service provider may lose customer trust.

For Canadian maintenance teams, calibration labs, utilities, contractors, manufacturers and quality managers, the key question is:

What decisions were made using this instrument while it may have been out of tolerance?

What Does Out of Tolerance Mean?

An instrument is out of tolerance when its measured error is outside the allowed limit for the calibration.

For example:

• A pressure gauge may read too high or too low.
• A thermometer may drift outside its acceptable error.
• A torque wrench may apply more or less torque than expected.
• A multimeter may fail at one part of its range.
• A transmitter may produce the wrong output signal for a given input.
• An RF wattmeter may have more uncertainty than the job can tolerate.

Calibration compares measuring equipment against a reference standard and documents how close the instrument is to the expected value. The source article defines calibration as a comparison against a higher-accuracy standard to detect, correlate, adjust, rectify and document instrument accuracy.

Why Out-of-Tolerance Results Matter

Out-of-tolerance results matter because instruments are used to make decisions.

A technician may use a calibrated instrument to decide whether to:

• Accept or reject a product
• Pass or fail a pressure test
• Approve a torque value
• Adjust a transmitter
• Set up a process
• Troubleshoot electrical equipment
• Confirm safety limits
• Sign off on customer work
• Release a system back into service

If the instrument was wrong, the decision may have been wrong too.

That does not automatically mean all past work failed. It means the work should be reviewed based on the instrument, the amount of error, the direction of error, the job tolerance and the risk of the application.

Calibration Range vs Instrument Range

One of the most important points in the source article is the difference between instrument range and calibration range.

The instrument range is what the instrument is capable of measuring.

The calibration range is the range over which the instrument is actually calibrated and used.

The source article gives the example of a pressure transmitter with a nameplate instrument range of 0 to 750 psig and a 4 to 20 mA output. If the engineer decides it will be calibrated only for 0 to 300 psig = 4 to 20 mA, then the calibration range is 0 to 300 psig, not the full 0 to 750 psig capability.

This distinction matters because an instrument may be capable of a wider range than the range used in the process.

A certificate should make clear what range was calibrated.

What Are Zero and Span?

In calibration, zero is the lower end of the calibrated range.

Span is the difference between the upper and lower range values.

Using the pressure transmitter example:

• Lower input value: 0 psig
• Upper input value: 300 psig
• Input span: 300 psig
• Output range: 4 to 20 mA
• Output span: 16 mA

The source article explains this exact example and warns not to confuse the range the instrument is capable of with the range for which it has been calibrated.

For Canadian calibration copy, this is useful because many disputes come from unclear range language.

A customer may assume the full instrument range was calibrated when only the process range was checked.

Accuracy, Tolerance and Uncertainty Are Not the Same Thing

These terms are often used casually, but they do not mean the same thing.

Accuracy

Accuracy describes how close a measurement is to the true or accepted value.

Tolerance

Tolerance is the permitted deviation from a specified value.

The source article quotes ISA dictionary-style definitions and recommends specifying tolerance in actual measurement units where possible, because this reduces mistakes caused by percentage calculations.

Uncertainty

Uncertainty describes the doubt associated with a measurement result.

No test instrument gives a perfectly exact reading. Every measurement has uncertainty.

NRC Canada’s calibration policy says traceability requires the ability to relate individual measurement results, with a stated uncertainty, through an unbroken chain of comparisons to a stated reference source, usually Canadian national measurement standards maintained by NRC or accepted intrinsic standards.

Why “In Tolerance” Does Not Mean “Perfect”

An instrument can pass calibration and still have measurement uncertainty.

That is normal.

A pass result means the instrument met the specified tolerance under the calibration conditions.

It does not mean the instrument has zero error.

The source article uses RF wattmeter examples to show how a reading that appears precise may still have a large possible measurement range once the test equipment’s tolerance is considered. In one example, a 14.9 W reading with a ±5% full-scale element could mean the actual transmitter output is anywhere from 12.4 W to 17.4 W.

The practical lesson is simple:

Know the capability of the instrument before trusting the reading.

Full Scale vs Reading-Based Tolerance

A common calibration mistake is misunderstanding whether tolerance is based on:

• Full scale
• Span
• Reading
• Count or digit
• Fixed engineering units
• A combination of terms

For example, ±5% of full scale is very different from ±5% of reading.

If a 100-unit full-scale device has ±5% of full-scale tolerance, the uncertainty is ±5 units at any reading.

At a reading of 15 units, that ±5 units is a large percentage of the actual reading.

This is why using the smallest suitable range often matters.

A test instrument may technically work, but it may not be accurate enough for the task.

Tolerances Should Be Written Clearly

The source article recommends specifying tolerance in actual measurement units whenever possible because it helps reduce confusion caused by percent-of-span or percent-of-reading calculations.

Clear tolerance examples:

• ±1 psi
• ±0.5°C
• ±0.02 mm
• ±2 ft-lb
• ±0.1 mA
• ±0.5% of reading
• ±0.25% of span

Poor tolerance examples:

• “Within spec”
• “Good enough”
• “Standard tolerance”
• “Manufacturer tolerance” without saying which specification
• “Accurate” without defining acceptance criteria

For critical instruments, define the tolerance before calibration begins.

Why As-Found Data Matters

As-found data shows how the instrument performed when it arrived for calibration, before adjustment or repair.

This is the most important data for out-of-tolerance risk review.

If the instrument fails as found, the team can ask:

• How far out of tolerance was it?
• Was the error high or low?
• Was the error present across the whole range or only at one point?
• What work was performed since the last calibration?
• Was the instrument used for safety, quality or customer acceptance?
• Were any readings close to pass/fail limits?
• Should recent work be reviewed?
• Should the calibration interval be shortened?

JM Test Canada’s calibration services page says calibration documentation includes as-found and as-left data, testing details and results on each piece of equipment.

Why As-Left Data Matters

As-left data shows how the instrument performed after adjustment, repair or final calibration.

This helps confirm whether the instrument is suitable to return to service.

A useful calibration record may show:

• As-found condition
• Adjustment or repair performed
• As-left condition
• Pass/fail status
• Tolerance applied
• Measurement uncertainty
• Standards used
• Traceability information

Without as-found and as-left data, a certificate may not provide enough information for quality review.

What to Do When Equipment Is Found Out of Tolerance

When an instrument is found out of tolerance, do not stop at the sticker.

Follow a structured review.

1. Review the Certificate

Check:

• Failed points
• Amount of error
• Direction of error
• Tolerance applied
• Range calibrated
• Measurement uncertainty
• As-found results
• As-left results
• Notes from the lab
• Whether adjustment or repair was performed

2. Identify Where the Instrument Was Used

Review:

• Job records
• Work orders
• Test reports
• Calibration logs
• Production records
• Customer records
• Maintenance records
• Pressure test reports
• Torque records
• Electrical test reports

3. Assess Risk

Ask:

• Was the instrument used for critical work?
• Were results close to acceptance limits?
• Could the error have changed the decision?
• Was the error in a direction that creates risk?
• Did the instrument affect product release, safety or compliance?
• Should the customer be notified?

4. Decide Corrective Action

Possible actions include:

• No action beyond documentation
• Review of affected records
• Re-testing
• Product hold
• Work recall
• Customer notification
• Process correction
• Calibration interval reduction
• Instrument repair or replacement
• Change in instrument selection

The source article makes the point that out-of-tolerance conditions can create business risk and may require special action, including review of previous work.

Not Every Out-of-Tolerance Result Has the Same Risk

A small out-of-tolerance result on a non-critical instrument may have limited impact.

A small out-of-tolerance result on a critical instrument may matter a lot.

Risk depends on:

• Instrument use
• Process criticality
• Direction of error
• Amount of error
• Measurement uncertainty
• Tolerance margin
• Customer requirements
• Safety impact
• Regulatory impact
• Whether the instrument was used near decision limits
• How often it was used
• How long it may have been out of tolerance

A failed calibration result should trigger review, not panic.

The right response depends on the actual risk.

Selecting the Right Test Equipment

The source article includes a practical example where a customer saw differences between transmit and receive levels on a communications test box. The issue was not necessarily that the unit failed calibration. The receiver and generator specifications were each broad enough that a large difference could exist while the unit still remained within manufacturer tolerance.

That example is important.

Sometimes the problem is not a failed instrument.

Sometimes the instrument is simply not accurate enough for the measurement being attempted.

Before selecting test equipment, confirm:

• Required measurement range
• Required tolerance
• Required uncertainty
• Required resolution
• Required traceability
• Environmental conditions
• Customer specification
• Industry requirement
• Whether accredited calibration is required
• Whether the instrument’s capability is suitable for the job

A calibrated instrument can still be the wrong instrument.

Canadian Traceability Context

For Canadian calibration work, traceability should be clearly documented.

NRC’s Metrology Research Centre is Canada’s national metrology institute and is responsible for developing and disseminating accurate measurement standards and providing metrology services. NRC says it provides accurate, traceable measurements that define the national standard and are recognized internationally.

A strong traceability record should include:

• Reference standard used
• Assigned value
• Stated uncertainty
• Calibration date
• Calibration provider
• Environmental conditions where relevant
• Unbroken chain to NRC, NIST or another recognized national metrology institute
• Certificate or report number

NRC’s calibration policy says a calibration result should include assigned value, stated uncertainty, the identity of the standards used and relevant environmental conditions where correction factors may apply.

ISO/IEC 17025 and Accreditation Scope

ISO/IEC 17025 is the international standard for testing and calibration laboratories. ISO describes it as setting requirements for laboratory competence, impartiality and consistent operation, helping ensure the accuracy and reliability of testing and calibration results.

However, a general ISO/IEC 17025 statement is not enough.

For Canadian teams, confirm:

• Is accredited calibration required?
• Is the exact instrument type covered?
• Is the exact range covered?
• Is the method covered?
• Is the required uncertainty covered?
• Is the lab’s scope current?
• Does the certificate identify the accreditation and scope where applicable?

NRC states that SCC is now fully responsible for assessment and accreditation to ISO/IEC 17025 for calibration and testing laboratories in Canada.

Calibration Interval and Out-of-Tolerance Risk

Out-of-tolerance results can help improve calibration intervals.

If an instrument repeatedly passes with plenty of margin, the interval may be appropriate.

If an instrument often fails as found, the interval may be too long or the instrument may be unsuitable for the application.

Calibration interval decisions should consider:

• Instrument history
• Previous as-found results
• Frequency of use
• Criticality
• Environmental exposure
• Handling conditions
• Drift pattern
• Manufacturer guidance
• Customer requirements
• Quality system requirements
• Cost of failure
• Replacement cost
• Risk of rework or recall

A fixed annual interval is common, but it is not automatically right for every instrument.

Calibration Certificate Checklist

A useful calibration certificate should include:

• Customer name
• Instrument manufacturer
• Model number
• Serial number
• Asset ID
• Calibration date
• Calibration due date, if assigned
• Calibration range
• Test points
• Tolerance applied
• As-found results
• As-left results
• Measurement uncertainty
• Reference standards used
• Traceability statement
• Environmental conditions where relevant
• Pass/fail result
• Adjustment or repair notes
• Calibration procedure or method
• Technician or lab identification
• Accreditation information where applicable

If the certificate does not show the tolerance used, the result may be hard to interpret.

If it does not include as-found data, the out-of-tolerance risk review may be incomplete.

Common Mistakes That Increase Out-of-Tolerance Risk

Confusing Instrument Range with Calibration Range

An instrument may be capable of a wider range than the range actually calibrated. The source article warns against confusing these two values.

Using Percent Tolerance Without Understanding the Base

Percent of full scale, percent of reading and percent of span can produce very different results.

Assuming a Digital Reading Is Exact

A digital display can look precise, but it still has uncertainty.

Ignoring As-Found Data

As-found data is what tells you whether past work may need review.

Treating Pass/Fail as the Whole Story

A pass/fail statement is useful, but it may not explain margin, drift, uncertainty or risk.

Using Test Equipment That Is Not Accurate Enough

A calibrated instrument can still be too inaccurate for the task.

Not Reviewing Failed Points

A device may fail at one point but pass elsewhere. The failed point must be compared to how the instrument is actually used.

Copying US-Only Traceability Language to Canada

For Canadian pages, do not rely only on “NIST traceable.” Canadian traceability language should also recognize NRC Canada and other recognized national metrology institutes where applicable.

JM Test Systems Canada Calibration Support

JM Test Canada’s calibration services page states that JM Test, formerly BHD, provides precision test, measurement and calibration solutions for industries across Canada. The page also refers to onsite and mobile calibration services, traceable certificates, calibration stickers, QR-coded certificates, as-found and as-left data, electronic records and retest notices.

For this Canadian article, avoid copying US-only language from the source page such as Baton Rouge contact details, US-only accreditation claims or blanket NIST-only traceability language.

Before publishing firm service claims, confirm:

• Current Canadian service availability
• Instrument type
• Measurement range
• Certificate type
• As-found/as-left data
• ISO/IEC 17025 scope
• Traceability documentation
• Onsite or mobile calibration availability
• Turnaround time
• Shipping or pickup options
• Customer portal access

Practical Takeaway

Out-of-tolerance results are not just calibration paperwork.

They are risk signals.

The original JM Test article explains that out-of-tolerance instruments can create business risk, customer trust issues and possible rework or recall, and it stresses the importance of understanding calibration range, tolerance, accuracy and measurement uncertainty.

For Canadian teams, the best approach is to:

• Define the tolerance before calibration
• Use the correct calibration range
• Understand measurement uncertainty
• Review as-found and as-left data
• Confirm traceability to NRC, NIST or another recognized national metrology institute
• Check ISO/IEC 17025 scope where accredited calibration is required
• Review past work when an instrument fails as found
• Match the test equipment capability to the job requirement

JM Test Systems Canada can support calibration needs where available with traceable certificates, as-found/as-left data, onsite and mobile service options and calibration management support. Confirm the exact range, method, certificate type, accreditation scope and Canadian service details before booking.