A laboratory can produce a beautifully formatted result that is wrong because the balance, thermometer, pipette or pressure instrument was unsuitable for the actual task. The Gold Standard of Laboratory Equipment Calibration Services is not sticking a new date label on every instrument. It is establishing what measurements matter, comparing equipment with suitable traceable references, reporting error and uncertainty honestly, and enabling the laboratory to decide whether equipment is fit for its intended use. Calibration provides measurement evidence; it does not by itself guarantee every later experiment will be correct.
Imagine a fictional Singapore food-testing laboratory using a precision balance, several pipettes and temperature loggers. A calibration supplier visits, prints three generic ‘passed’ labels and departs. Weeks later, a quality manager discovers that the balance was checked only at a mass far above the laboratory’s routine sample size, while a pipette was tested outside the range most analysts use. Certificates exist, but they do not support the real results. A competent service would agree critical ranges and tolerances, document measurement conditions, and return specific readings, uncertainties and relevant limits for the laboratory’s own decisions.
The Singapore Accreditation Council (SAC) describes SAC-SINGLAS as the national laboratory accreditation scheme using ISO/IEC 17025:2017 for relevant testing and calibration, with accreditation granted for specific fields and methods, not as a blanket endorsement of every service a company offers. SAC’s August 2026 document catalogue lists measurement traceability, uncertainty, calibration capability and equipment quality-assurance guidance. The essential principle is simple: a measurement certificate must show what can actually be trusted, within which stated limits.
Quick answer: what makes a laboratory calibration trustworthy?
It covers the instrument’s real working range, compares it with suitable documented traceable references, reports relevant results and measurement uncertainty, and supports an explicit fitness-for-purpose decision. SAC-SINGLAS accreditation applies to specific calibrated quantities, methods and capabilities, not every service offered by an accredited company. A date label without actual data is not a complete result.
Follow the instrument from measurement purpose to qualified use
Use the chapters to distinguish calibration, adjustment and verification; traceability, uncertainty and tolerances; device-specific procedures; nonconforming equipment; and the appropriate handover.
1. Start with the laboratory’s actual decision
Identify what the measurement is used to decide
A school science lab, contract-testing laboratory, pharmaceutical facility and manufacturing quality laboratory may all use balances and thermometers, but their decisions and critical tolerances differ. Ask what result the device supports, who uses it, and what error would change the decision. A calibration that ignores the intended measurement is a paperwork exercise.
Describe fitness for use separately from the provider’s work
The customer should set or approve relevant acceptance requirements, and the calibration supplier should demonstrate the instrument’s actual measurement performance. Where the tolerance is defined by regulation, validated method or client specification, cite the correct source. A calibration lab cannot invent a broad ‘industry standard pass’ when no applicable criterion has been agreed.
Back to chapters · Next chapter
2. Inventory each device before booking the visit
Keep exact model, range and identity
Record manufacturer, model, serial or permanent asset ID, capacity, resolution, relevant accessories, location and calibration history. Two pipettes of the same brand may differ in range or condition, so an invoice saying ‘five pipettes calibrated’ may not support the specific results an auditor needs.
Link records to the actual user location
The report should allow the food lab to connect each certificate to the balance or instrument used for a test batch. Where equipment can move between rooms, preserve the identity and relevant configuration rather than treating a room number as the only identifier.
Back to chapters · Next chapter
3. Distinguish calibration, adjustment and verification
Understand calibration as a relationship between indications and references
Calibration establishes information about an instrument’s indications relative to suitable standards under known conditions. It may reveal bias or variation, even if the equipment remains usable. Adjustment changes how an instrument behaves; verification assesses conformity to agreed requirements. These are related but not identical activities.
Never let a ‘passed’ sticker substitute for data
A device can have a valid calibration certificate with significant error, and that certificate may still be technically useful. Whether it passes the laboratory’s own tolerance may need a separate conformity decision. A service should identify which work and judgment it performed and what remains with the customer.
Back to chapters · Next chapter
4. Understand scope-specific SAC accreditation
Check the actual accredited measurement capability
SAC-SINGLAS accreditation recognises a laboratory for defined fields, methods, ranges and capabilities. Accreditation as a calibration laboratory does not imply that every model, range or unit offered in a quotation is covered. Verify the relevant scope in the official directory where an accredited result is requested.
Do not use accreditation marks beyond their permitted scope
SAC’s 2026 terms and marks rules govern how accredited work is presented. An accredited company may also carry out non-accredited jobs, which should be identified honestly. A customer requesting a specific accredited pipette or mass calibration should receive the appropriate report for that scope, not a generic certificate with a logo.
Back to chapters · Next chapter
5. Use ISO/IEC 17025 correctly
Recognise technical competence and a working quality system
ISO/IEC 17025:2017 addresses competent testing and calibration laboratories, including methods, resources, impartiality, measurement evidence and management processes. It is not a magic tolerance for every instrument. The relevant laboratory, method and scope determine what must be done to establish trustworthy results.
Distinguish it from medical laboratory accreditation
SAC identifies ISO 15189:2022 for medical laboratory accreditation and ISO/IEC 17025 for testing and calibration scope. A clinical laboratory’s own governance may therefore differ from a commercial dimensional metrology lab. Use the standard appropriate to the actual organisation and service.
Back to chapters · Next chapter
6. Make traceability a real unbroken chain
Link results through calibrated references
Metrological traceability means that a measurement result can be related to a reference through a documented unbroken chain of calibrations, each contributing to uncertainty. SAC-SINGLAS 006 describes the principle. Merely writing ‘traceable to NIST’ or ‘traceable to SI’ without supporting evidence and uncertainty is inadequate.
Identify the relevant reference standard and its status
Check the reference’s identity, suitable range, calibration history, and whether it remained within relevant validity and control conditions when used. A reference certificate is necessary evidence, but its availability alone does not prove that the supplier applied the correct method to the client’s instrument.
Back to chapters · Next chapter
7. Specify the intended range and calibration points
Cover where the laboratory really measures
A balance that weighs tiny portions may require evidence in its low working range, not only at its maximum capacity. A pipette used repeatedly near the lower setting may need information there as well as at the upper setting. Agree representative points based on the customer’s method and manufacturer specifications.
Record what was not covered
If the supplier’s method or capability excludes part of the range, state the limitation clearly. A certificate for one calibrated range should not be used to assert validity over every possible scale graduation or operating condition.
Back to chapters · Next chapter
8. Match tolerance to the actual process
Separate manufacturer specification from customer acceptance
Manufacturer claims, accredited method requirements, client tolerances and regulatory limits can differ. A device may be within the maker’s specification but still inadequate for a demanding analytical method. The laboratory needs a defensible fit-for-purpose criterion rather than selecting whichever tolerance yields a favourable ‘pass’.
Document who approved the decision rule
Where an accredited report states conformity, the uncertainty and decision rule may affect the conclusion. Define and communicate the decision rule under the appropriate ISO/IEC 17025 arrangements. The supplier should not quietly apply an unfamiliar guard band and label a borderline instrument compliant without explanation.
Back to chapters · Next chapter
9. Estimate and report measurement uncertainty
Do not confuse an error with uncertainty
Error is the difference between an indication and a reference under defined conditions; measurement uncertainty expresses a quantified doubt about the result. They are not synonyms. Both can matter when the laboratory decides whether to accept an instrument.
Use a meaningful, method-based uncertainty evaluation
SAC’s March 2026 technical guide addresses evaluation and expression of measurement uncertainty. The certificate should describe relevant uncertainty in suitable terms and explain coverage factor or confidence convention where used. A single optimistic ‘accuracy ±0’ claim is a warning, not evidence of exceptionally careful measurement.
Back to chapters · Next chapter
10. Explain a borderline result with a real example
Use a result near the tolerance boundary
Suppose a laboratory’s allowable indication error is within ±0.10 g for a particular application. A calibration finds +0.08 g at a critical point with expanded measurement uncertainty of 0.04 g. The observed error lies inside the stated limit, but uncertainty can affect a conformity conclusion under the selected decision rule. The report should not silently call it a universally safe pass.
Keep numerical conclusions tied to the agreed rule
Whether the instrument is acceptable depends on the actual method, tolerance, uncertainty and risk approach. A professional explains the decision and the consequences for the laboratory’s measurements, instead of hiding an ambiguous result behind green and red labels.
Back to chapters · Next chapter
11. Control the environmental conditions of calibration
Record temperature, humidity and other influences
Some instruments respond to room temperature, humidity, vibration, airflow, electromagnetic effects or power quality. A bench-top balance near a vent may behave differently from one on a suitable stable surface. The provider should know which environmental conditions affect its method and document them where relevant.
Distinguish on-site and laboratory conditions
An in-house calibration room may permit tighter control than a busy factory floor. On-site work can be appropriate, but its uncertainty and practical limits need evaluation. Do not issue the same unquestioned result from a vibrating loading dock as from a properly controlled laboratory when conditions materially differ.
Back to chapters · Next chapter
12. Use suitable methods for balances
Check more than one nominal mass when needed
Mass-measurement performance can involve indication, repeatability, eccentricity and the actual range of use. The relevant method determines which checks are required. A service testing one heavy mass in the centre of the pan may miss a problem important to routine small sample measurements.
Preserve the actual working configuration
Record whether accessories, draught shields, calibration weights, vibration conditions or software settings differ from the way the customer uses the balance. Do not adjust a unit and then return it to a different configuration without documenting what changed.
Back to chapters · Next chapter
13. Calibrate pipettes for how analysts actually dispense
Understand volume setting and tip technique
Air-displacement and other pipettes can vary with settings, use and technique. The service should follow an appropriate method and relevant equipment conditions, measuring at points meaningful to actual use. A single nominal volume claim does not cover the entire working range.
Separate the instrument from operator performance
Pipette calibration and a technician’s handling technique interact, but an analyst’s day-to-day pipetting errors may also arise from training, tips and environmental conditions. The supplier should provide usable results and relevant method information without implying its certificate proves every later person will pipette accurately.
Back to chapters · Next chapter
14. Assess temperature instruments across their application
Match the instrument to the critical temperatures
A probe used for refrigeration, an incubator monitor and an autoclave reference operate at different temperature ranges and under different conditions. Define where the laboratory must trust the reading. Checking a thermometer only near room temperature cannot support every freezer or high-temperature use.
Treat sensor response and placement seriously
Temperature readings can lag changes and depend on probe positioning, contact and stability. The calibration method should address appropriate equilibrium and documented measurement conditions, while the user must separately validate how the sensor behaves in its actual equipment.
Back to chapters · Next chapter
15. Understand the difference between calibration and mapping
Mapping measures variation within a space
A cold room or incubator can contain temperature differences between shelves and locations. Calibrating the logger establishes information about that instrument; mapping examines the environmental distribution under defined conditions. These are different services, and one cannot automatically replace the other.
Use the right outcome for temperature-controlled goods
If a customer needs evidence that an entire refrigerator holds the required range, a single calibrated probe may be necessary but not sufficient. Plan the correct temperature mapping, qualification or routine monitoring with the relevant specialists.
Back to chapters · Next chapter
16. Treat pressure, flow and humidity devices as separate cases
Choose the appropriate reference quantity
Pressure gauges, flow meters and humidity probes have different measurement principles and reference requirements. A provider should identify the actual range, medium, environmental constraints and safety conditions before work. One generic calibration method cannot supply reliable results for all measurement types.
Report the configuration used during calibration
Installation orientation, connections, hoses, fittings or operating medium may affect a result. Document what matters for future comparison and safe reuse, while avoiding unauthorised intrusion into pressure vessels or regulated systems.
Back to chapters · Next chapter
17. Separate measurement instrument care from repair
Decide whether a device needs cleaning, adjustment or replacement
An instrument may fail because of dirt, wear, physical damage or drift. Calibration reveals information but may not repair the cause. If servicing or adjustment is needed, explain the scope, authorisation and effect on previous measurement evidence. Do not repair secretly and issue only a post-adjustment certificate.
Give the laboratory both as-found and as-left information where relevant
When adjustment changes performance, initial results may be critical to assessing earlier laboratory work. A supplier should preserve and report appropriate as-found conditions and describe what changed, followed by suitable re-calibration evidence.
Back to chapters · Next chapter
18. Manage devices that fail acceptance
Quarantine or restrict use proportionately
Where a laboratory’s acceptance rule is not met, mark and control the instrument under the client’s quality system. A failed result may require repair, reassignment to a less demanding task or withdrawal. A vendor should not automatically discard an expensive device or declare every failure harmless.
Assign the next decision to the right owner
The calibration laboratory supplies measurement evidence; the customer decides whether affected testing must be repeated, past batches investigated or products held. A responsible supplier highlights meaningful out-of-tolerance findings promptly instead of burying them in a later PDF.
Back to chapters · Next chapter
19. Investigate the effect on previous measurements
Think backwards from the last reliable point
If a device is found outside tolerance, the customer may need to assess when the deviation developed and whether earlier results were materially affected. Review the instrument’s prior calibrations, internal check records, methods and the significance of possible error. No single calibration certificate tells the exact date drift began.
Avoid promising retrospective certainty without evidence
Where uncertainty remains, disclose it and choose a risk-based investigation. Correcting the instrument now does not retroactively make yesterday’s sample results accurate.
Back to chapters · Next chapter
20. Choose calibration intervals using risk and history
Do not confuse the next due date with a universal law
The right interval depends on device use, criticality, stability, environment, manufacturer recommendations, regulatory or method requirements and previous performance. ISO/IEC 17025 does not set one single annual calibration date for every device. Laboratories should justify their own schedules.
Change the programme when evidence warrants
Recurring drift may call for shorter intervals or better intermediate checks, while stable equipment can support a defensible programme within applicable requirements. A calendar-based subscription is useful only if it reflects the actual measurement risk.
Back to chapters · Next chapter
21. Use intermediate checks between formal calibrations
Detect problems before the next annual certificate
A laboratory may employ suitable check standards, control charts, daily balance checks or comparison measurements to detect drift. These activities complement, rather than necessarily replace, required external calibration. Define limits and escalation so a failed internal check is not casually ignored.
Keep simple records meaningful
Document which reference was used, the result, person and action. A daily check box signed without a real measurement can provide false confidence, particularly when the instrument supports consequential release decisions.
Back to chapters · Next chapter
22. Consider traceability across outsourced services
Know where the measurement was actually made
A calibration provider may subcontract specialist work. The customer should be told relevant arrangements and whether each result is covered by the claimed accreditation scope. The certificate must identify the responsible laboratory and allow the measurement record to remain traceable.
Avoid purchasing by accreditation logo alone
Verify the relevant service capability, turnaround, uncertainty, location and terms. A very low price can be misleading if important ranges or accredited reporting are unavailable.
Back to chapters · Next chapter
23. Keep calibration records tamper-evident and retrievable
Make the certificate specific to the measured instrument
An effective record identifies equipment, result, method, references, relevant conditions, uncertainty, date and any conformity statement with the applicable decision rule. Review it for missing units, ambiguous serials or reported points not matching the order. Electronic certificates need reliable integrity and access controls.
Protect the distinction between original results and copies
If a certificate is corrected after issue, document the revision rather than silently overwrite the original evidence. Laboratories need to understand what version governed a past method or audit decision.
Back to chapters · Next chapter
24. Avoid using a calibration label as a complete report
Explain what the label can legitimately convey
A sticker can help users identify asset status and due date, but it is an operational cue. It cannot communicate all measured errors, uncertainty, limited scope or conditions. A valid-looking sticker may even remain on a device whose calibration result is unacceptable for one specific application.
Teach users where to find the real certificate
Make records easy for analysts, supervisors and quality staff to retrieve. The most valuable service outcome is an informed equipment decision, not a tidy instrument case with matching labels.
Back to chapters · Next chapter
25. Plan site access, contamination and worker safety
Prepare equipment and workspace for competent testing
Laboratory work can involve chemicals, biological hazards, electricity, heat, pressurised equipment or moving machinery. Agree decontamination requirements, safe instrument access, electrical isolation and practical workstation conditions. A calibration engineer should not handle contaminated apparatus simply because an appointment is scheduled.
Preserve the laboratory’s work and samples
Coordinate downtime and movement without losing custody of instruments or interfering with ongoing experiments. If equipment must leave the site, document collection, condition, protective packing and return.
Back to chapters · Next chapter
26. Make the quotation cover the actual capability
Specify parameters, range and report status
Quote each instrument, calibration points, required uncertainty, accredited or non-accredited report, on-site costs, adjustment, repairs, delivery, urgent service and applicable re-tests. This lets customers compare equivalent coverage rather than the count of certificates printed.
Communicate charges transparently
Singapore’s CCS price-transparency guidance encourages disclosure of unavoidable fees. A cheap ‘instrument calibrated’ rate may be misleading if the actual working range or accreditation requirement carries an unmentioned compulsory surcharge.
Back to chapters · Next chapter
27. Treat delays and failures as quality events
Report when a reference or test method becomes unavailable
Unexpected faults, equipment contamination, lack of stable conditions or expired reference standards can prevent a planned calibration. Stop rather than issue a result unsupported by valid references or method conditions. Tell the client which measurements are delayed and what alternatives are legitimate.
Keep a real exception record
An unresolved instrument should not appear as calibrated merely to meet an internal target. Show what was attempted, why the process stopped and who owns the next competent action.
Back to chapters · Next chapter
28. Provide an acceptance test the user can understand
Connect the certificate to actual laboratory work
For the food laboratory, the QA manager checks that the balance, pipette and logger identifiers match the equipment and that results cover their critical working ranges. The reported errors, uncertainty and any conformity claims are checked against method tolerances. Where a device fails, the laboratory updates its status and assesses relevant prior results.
Close only what has been proven
A service can be complete when accurate evidence reveals a failed instrument and the laboratory has an actionable plan. It should not force every device into an artificial ‘pass’ category for customer satisfaction.
Back to chapters · Next chapter
29. Measure competence by decisions improved
Use outcomes instead of certificate counts
Useful indicators include missed critical ranges, traceability gaps, invalid reference dates, repeated out-of-tolerance results, delayed reporting, transcription corrections, unsafe handling and user understanding of uncertainty. Large numbers of labels or low turnaround averages do not alone establish measurement quality.
Make reliable science the ultimate service outcome
The gold standard is a laboratory that knows which instruments support which decisions and can defend their measurement evidence. Honest uncertainty, correct scope, technical competence and prompt follow-up protect scientific conclusions far more effectively than a generic certificate proclaiming everything perfect.
Worked example: a balance, pipette and temperature logger
The fictional food-testing laboratory regularly weighs small samples, dispenses working volumes near the lower end of one pipette range and logs chilled product temperatures. A weak provider would supply generic ‘passed’ labels based on one convenient point per instrument. A competent provider agrees the critical ranges and appropriate methods, verifies the reference chain and reports actual errors, uncertainty, and any adjustments. The laboratory applies its own documented fitness and method requirements instead of accepting the label as the final decision.
If one pipette is outside an agreed limit, its as-found result remains important even after adjustment. The QA manager may need to check earlier laboratory work and impose appropriate controls until the device is accepted for use. The calibration business helps by presenting accurate, specific evidence, not by silently converting an inconvenient failure into a green tick.
Laboratory calibration acceptance matrix
| What the laboratory needs | Evidence during the service | Handover that actually supports quality |
|---|---|---|
| Correct instrument and range | Asset identity, use and critical points agreed | Specific reported data over the intended range |
| Traceable comparison | Suitable current references and calibration chain | Defensible traceability and uncertainty |
| Honest conformity | Acceptance criteria and decision rule defined | No unsupported ‘pass’ or accreditation claim |
| Changes and faults visible | As-found/as-left results and adjustment history recorded | Proper nonconforming-device decisions |
| Sustained measurement control | Intervals, intermediate checks and retrievable certificate | Users know what remains fit for purpose |
Frequently asked calibration questions
Does ISO/IEC 17025 accreditation cover every test or calibration a supplier sells? No. Singapore Accreditation Council confirms that accreditation covers specific fields, methods and capabilities. Check the actual scope.
Are calibration and adjustment the same? No. Calibration establishes measurement relationships; adjustment changes an instrument; verification evaluates conformity to an agreed requirement. They may be combined but should be identified.
Does a calibration certificate automatically mean the instrument passed? No. A certificate can legitimately show an error that exceeds the customer’s fit-for-purpose tolerance. Conformity requires an appropriate specification and decision rule.
Is every laboratory instrument legally required to be calibrated once a year? No universal ISO/IEC 17025 annual rule applies. Intervals should reflect use, stability, method or regulatory requirements and monitoring history.
Singapore SAC-SINGLAS and international measurement sources
Authoritative sources: Singapore Accreditation Council: SAC-SINGLAS Accredited Testing and Calibration Laboratories (August 2026); SAC Laboratory Documents: ISO/IEC 17025, Traceability, Calibration Capability and March 2026 Uncertainty Guidance; SAC Accreditation FAQ and Scope Boundaries; NIST SI Units and Metrology References; Singapore Consumer Price Transparency Guidelines.
Connected eduKate Gold Standard services: Indoor Air Quality Testing Services; Water Tank Cleaning Services; Refrigerated Transport Services; Records Management Services.