Anchor testing is frequently requested late in a construction programme, often when a contractor needs evidence that installed fixings are suitable to remain in service. The difficulty is that terms such as proof testing, pull-out testing, preliminary testing and destructive testing are regularly used as though they mean the same thing.
They do not. For UK construction, the current reference is BS 8539:2012+A1:2021 — Code of practice for the selection and installation of post-installed anchors in concrete and masonry. The standard covers much more than testing: anchor selection, design, installation, supervision, inspection and testing all form part of the process.
For contractors, the most important point is that an anchor test should not begin with somebody arriving on site and deciding how hard to pull. The anchor type, substrate, test purpose, test load, sampling regime and acceptance basis should already be understood before the testing equipment is connected.
The key message: anchor proof testing is primarily a quality-control check on installed anchors. It does not determine an anchor's ultimate capacity, it does not replace anchor design, and the testing contractor should not invent the proof load or sampling regime on site.
Jump to: What BS 8539 covers | Test types | How many anchors | Test loads | Testing procedure | Pass or fail | Why anchors fail | UKAS | Contractor checklist | FAQ
What Is BS 8539 and Why Does It Matter?
BS 8539:2012+A1:2021 is the current UK code of practice dealing with the selection and installation of post-installed anchors in concrete and masonry. It applies to post-installed anchors fitted into drilled holes and covers both mechanical and bonded anchor systems within its scope. On a construction project, this can include fixings used for structural steelwork, façade systems, building services, temporary works, brackets, supports and many other applications. Testing is therefore only one part of the control process. A satisfactory test cannot correct an anchor that was wrongly selected, wrongly designed or installed contrary to the manufacturer's requirements.
| Stage | Typical Question | Why It Matters |
|---|---|---|
| Selection | Is this the correct anchor for the substrate, environment and application? | Testing cannot compensate for selecting the wrong fixing system. |
| Design | What actions must the fixing resist? | The design provides the engineering basis for the anchor and any test requirement. |
| Installation | Was drilling, cleaning, embedment, resin curing or tightening completed correctly? | Installation quality is one of the principal reasons proof testing is used. |
| Inspection | Does the installed fixing correspond with the approved arrangement? | Inspection identifies visible departures before testing or loading. |
| Testing | Does the selected anchor sustain the specified test load without unacceptable behaviour? | Testing provides measured site evidence, but only for the purpose and anchors covered by the test regime. |
Anchor Proof Testing vs Pull-Out Testing
The expression pull-out test is widely used on UK sites, but it is a broad description rather than a precise statement of test purpose. A contractor requesting a “pull test” should therefore confirm what the engineer actually wants the test to demonstrate.
Proof Testing
A proof test applies a predetermined tensile load to an installed production anchor. The anchor is normally intended to remain in service after the test. The purpose is principally to provide quality-control evidence about the installed fixing. It is not intended to pull the anchor to destruction or determine its maximum resistance.
Preliminary or Site-Suitability Testing
Where the actual substrate is uncertain, variable or not adequately represented by published anchor performance data, preliminary testing may be required before the production fixing regime is confirmed. These tests are normally carried out using sacrificial anchors in representative substrate. Their purpose is different from routine proof testing: they help establish how the proposed anchor behaves in the actual base material so that the designer can make an engineering decision.
Ultimate or Destructive Testing
A destructive test deliberately investigates resistance towards failure. The test anchor is sacrificial and should not be treated as an installed production anchor afterwards.
| Test | Main Purpose | Anchor Remains in Service? |
|---|---|---|
| Proof test | Quality-control check against a specified proof load. | Normally yes, provided the test is satisfactory. |
| Preliminary / suitability test | Establish site-specific performance where substrate behaviour requires investigation. | Normally no; test anchors are typically sacrificial. |
| Ultimate / destructive test | Investigate the failure resistance or failure mechanism. | No. |
Contractor warning: specifying “pull-out testing” without stating whether proof testing, preliminary testing or destructive testing is required can produce the wrong test and the wrong report.
How Many Anchors Need Testing Under BS 8539?
There is no universal rule saying that every project must test 5%, 10% or any other single percentage of installed anchors. BS 8539 Annex B contains sampling arrangements for proof testing, including combinations linking the number of production anchors tested with the proof-load factor. Commonly encountered arrangements include 1 in 25 anchors associated with a 1.25 proof-load factor and 1 in 40 anchors associated with a 1.5 proof-load factor in the circumstances covered by the standard.
Those numbers should not be copied onto every project without checking the applicable conditions. Anchor approval, supervision, substrate, installation conditions, structural redundancy, consequence of failure and the project specification can all affect the appropriate testing regime. Annex B also contains minimum test-number provisions, meaning that quoting only a percentage can be misleading on smaller fixing populations. For safety-critical or non-redundant installations, the designer or specifier may require a significantly higher level of testing, potentially including testing every applicable production anchor.
| Site Request | Problem | Better Approach |
|---|---|---|
| “Test 10% of the anchors.” | No technical basis for the percentage has been stated. | Ask the designer or specification author to define the test regime. |
| “Just test three anywhere.” | The selected anchors may not represent the installation population. | Define populations by anchor type, installer, substrate, area and installation conditions. |
| “One anchor passed, so the installation is approved.” | One result cannot automatically represent every fixing on the project. | Follow the agreed sampling and acceptance regime for the defined anchor population. |
Who Decides the Anchor Proof-Test Load?
The testing contractor should not arrive on site and invent the required test load. The proof-test requirement should come from the project design or specification, normally through the structural engineer, façade engineer, temporary works designer, anchor designer or other competent specifier responsible for the fixing application.
BS 8539 proof-test provisions relate the proof load to the relevant characteristic action. The important practical distinction is between an unfactored characteristic action and a factored design action used in structural calculations. Contractors should therefore avoid sending a drawing containing an unidentified load and asking the test technician simply to apply “1.5 times this figure”. If the figure is already a factored design value, multiplying it again can produce an inappropriate test load.
Equally, there should be no blanket attempt to convert a design action back to a characteristic action using an assumed universal divisor. The partial factors depend on the actual structural design and load combination. Where the correct basis is unclear, the responsible designer should issue the required proof load or confirm the calculation basis before testing starts.
Before Booking the Test
The safest instruction is not “pull to 15 kN because that is what we normally use”. It is: identify the anchor population, provide the engineer-approved proof load and sampling regime, and state the acceptance requirement in the test brief.
How Anchor Proof Testing Is Carried Out on Site
The detailed apparatus and procedure should follow the applicable standard, manufacturer requirements, project specification and testing method. In practical terms, a controlled site proof test normally follows the sequence below.
- Confirm the test brief. Verify the anchor population, fixing type, diameter, substrate, required proof load, sampling regime and acceptance basis.
- Identify the anchor. Confirm that the test position corresponds with the anchor or fixing identified by the project team.
- Inspect the surrounding substrate. Record visible cracking, damaged masonry, poor concrete, unusual edge conditions or other features that may affect testing.
- Check the testing equipment. The load equipment should be appropriate for the load range and supported by current calibration evidence.
- Set up the reaction arrangement. The tester must react sufficiently clear of the anchor so that the rig does not locally confine the substrate and artificially improve the result.
- Align the test load. Tensile load should be applied as axially as practicable. Significant eccentricity can introduce bending or distort the test result.
- Apply load in a controlled manner. Increase the tensile load smoothly to the specified proof load rather than shock-loading the fixing.
- Maintain and observe. Where the specified regime requires it, the proof load is held while the tester observes load stability, movement and substrate behaviour.
- Unload safely. Release the load in a controlled manner and check the anchor and surrounding material.
- Record the evidence. The report should link the result to the tested anchor location and include the load applied, relevant anchor information, equipment identification and test outcome.
A poor reaction arrangement can materially distort a test. If the legs of a pull-test bridge are positioned too close to the fixing, the rig can restrain the local failure zone rather than allowing the substrate to respond naturally. Poor alignment, incorrect load units, damaged base material or unsuitable edge conditions can also undermine the validity of the result. Testing is a loading operation and should be controlled accordingly. For wider guidance on site loading arrangements, see What a Structural Testing RAMS Must Contain Before Loading Starts.
What Counts as a Pass?
A proof test is not simply a question of whether the gauge reaches the specified number. The anchor should achieve the specified proof load and behave satisfactorily during the required test period. Progressive movement, slip, visible substrate cracking, spalling or other signs of distress require investigation and may make the result unacceptable.
Where a specified hold period applies, the anchor must remain stable through that period rather than reaching the target load momentarily and immediately being unloaded. Most importantly, a successful proof test does not prove the ultimate capacity of the anchor. It demonstrates that the tested fixing sustained the specified proof load under the particular site conditions at the time of testing.
| A Successful Proof Test Can Support | A Successful Proof Test Does Not Automatically Prove |
|---|---|
| That the tested anchor sustained the specified proof load. | The ultimate failure capacity of the anchor. |
| Quality-control evidence for the tested installation. | That every untested anchor on the project is satisfactory. |
| Evidence within an agreed representative sampling regime. | Long-term performance against corrosion, creep, fatigue, fire or other conditions not represented by the test. |
What Happens If an Anchor Fails?
A failed proof test should not automatically lead to the failed anchor simply being replaced and the testing team moving to the next location. The failure may indicate a local defect, an installation problem, unexpected substrate behaviour or a wider systematic issue affecting other anchors in the same population.
The affected installation should therefore be held while the project team identifies the likely cause and agrees the appropriate response with the responsible designer or specifier. Depending on the governing test regime and the significance of the failure, this can include increased sampling, further investigation, replacement of affected anchors or potentially testing the wider population.
Common Reasons Anchors Fail Site Testing
| Potential Cause | What Should Be Checked |
|---|---|
| Incorrect hole diameter or depth | Drill bit size, drilling method and required embedment. |
| Poor hole cleaning | Whether the manufacturer's required blowing and brushing sequence was followed. |
| Resin curing issue | Product, curing time, substrate temperature and whether loading occurred too early. |
| Incorrect anchor or resin | Installed product against the design specification and manufacturer documentation. |
| Weak or variable substrate | Brick, block, mortar, concrete condition and whether preliminary testing should be considered. |
| Water, dust or contamination | Hole condition and installation requirements for the selected anchor system. |
| Incorrect torque or installation procedure | Manufacturer torque requirements, installation tools and installer competence. |
| Edge distance or spacing issue | Installed geometry against the approved anchor design. |
Do not treat a failed test as a testing-company problem until the cause is understood. The result may be exposing an installation, design or substrate issue that existed before the test equipment was attached.
Does BS 8539 Require UKAS-Accredited Anchor Testing?
This is another area where contractual requirements and the British Standard are frequently confused.
BS 8539 does not itself create a universal statutory requirement that every anchor proof test must be carried out by a UKAS-accredited testing organisation. A project specification, client, principal contractor, infrastructure authority or other contract document may nevertheless require accredited testing.
It is also important to distinguish equipment calibration from organisational accreditation A calibration certificate provides evidence about the performance and traceability of the measurement equipment. UKAS accreditation, where held for the relevant activity, relates to an organisation's demonstrated competence for a defined accredited scope. One should not be presented as though it automatically proves the other.
Before appointing a testing provider, contractors should check the competence required by the project, relevant technician training and experience, equipment calibration, test methodology, reporting capability, insurance and any project-specific accreditation requirement.
Contractor Checklist Before Booking Anchor Testing
A large proportion of site delays can be avoided if the test brief is complete before mobilisation.
- Anchor specification: manufacturer, product, diameter and type.
- Substrate: concrete, brick, block, masonry or other relevant base material.
- Fixing population: how many anchors are installed and which anchors belong to the same test population.
- Test purpose: proof testing, preliminary testing or another defined requirement.
- Required proof load: obtain the value or approved basis from the responsible engineer or specifier.
- Sampling regime: confirm how many anchors must be tested and how they should be selected.
- Acceptance criteria: define what constitutes an acceptable result before testing.
- Anchor locations: drawings, grid references, elevations or marked test positions.
- Installation evidence: installer details, product records, curing information or other QA evidence where relevant.
- Access: make sure the testing rig can physically connect to the fixing and react safely against the substrate.
- Project requirements: check whether CFA certification, specific reporting, UKAS accreditation or another contractual condition applies.
- Reporting: agree what anchor identification, calibration information, photographs and results need to appear in the final report.
Need Anchor Testing?
For an accurate quotation, provide the anchor specification, substrate, number of anchors, test locations, required proof load, sampling regime and any project-specific reporting or accreditation requirements. If the engineering test load has not yet been defined, this should be confirmed by the responsible designer or specifier before site testing begins.
The Most Common Contractor Mistakes
| Mistake | Better Approach |
|---|---|
| Asking for a generic “pull-out test”. | State exactly what type of test and engineering question needs to be addressed. |
| Telling the tester to choose the load. | Obtain the proof load or calculation basis from the responsible designer. |
| Assuming 5% or 10% is always required. | Confirm the sampling regime against BS 8539, the design and the project specification. |
| Testing resin anchors before full cure. | Follow the manufacturer's curing requirements for the actual substrate temperature. |
| Treating a passing proof test as ultimate-capacity evidence. | State clearly that a proof test only demonstrates performance to the specified proof load. |
| Automatically replacing one failed fixing and continuing. | Investigate the failure and agree the revised testing or remediation regime. |
Evidence-Based Summary
BS 8539:2012+A1:2021 remains the current UK code of practice for the selection and installation of post-installed anchors in concrete and masonry.
Proof testing is primarily a quality-control check. It should not be confused with destructive testing used to establish failure resistance.
There is no universal 5% or 10% rule for every anchor installation. Sampling and proof load need to be considered together and applied to the appropriate project circumstances.
The test provider should not invent the engineering test load. The design or specification should establish the required basis before testing starts.
A successful proof test demonstrates performance to the specified proof load at the tested location. It does not determine ultimate capacity or automatically validate every other anchor on the project.
FAQ: BS 8539 Anchor Testing
What is the current version of BS 8539?
The current version identified in the research is BS 8539:2012+A1:2021, Code of practice for the selection and installation of post-installed anchors in concrete and masonry. Current project requirements should always be checked against the latest BSI publication status.
Is an anchor proof test the same as a pull-out test?
“Pull-out testing” is commonly used as a general site term for applying tensile load to an anchor. A proof test is more specific: it applies a defined proof load to an installed anchor without intentionally taking the fixing to failure.
How many anchors need to be tested?
There is no single percentage that applies to every project. BS 8539 contains sampling provisions, but the appropriate regime depends on the particular installation, approval and supervision conditions, structural significance and project specification. The responsible designer or specifier should define the requirement.
Who decides the anchor test load?
The engineering basis should normally come from the structural engineer, façade engineer, temporary works designer, anchor designer or other competent specifier responsible for the fixing. The testing contractor should not guess a proof load.
Does a passing proof test prove the anchor's ultimate capacity?
No. Proof testing is intended to demonstrate that the tested fixing sustains the specified proof load without unacceptable behaviour. It does not establish the ultimate failure resistance.
What happens if one anchor fails?
The failure should trigger investigation rather than simply replacing one fixing and continuing. The designer or specifier should consider the cause and determine whether increased sampling, remediation or wider testing is required.
Do resin anchors need to cure before testing?
Yes. Bonded anchors should not be proof loaded until the manufacturer's required curing period has been achieved for the actual installation and substrate temperature.
Does BS 8539 require UKAS-accredited testing?
BS 8539 does not itself impose a universal statutory requirement for every anchor proof test to be undertaken by a UKAS-accredited organisation. Accreditation may, however, be a contractual or project-specific requirement and should be confirmed before appointment.
Is calibrated equipment the same as UKAS-accredited testing?
No. Calibration provides evidence about the performance and traceability of measurement equipment. Organisational accreditation relates to competence for a defined accredited activity and scope. The two concepts should not be confused.
Can one successful test certify every anchor on the site?
No. A test result applies directly to the tested anchor. Representative sampling may provide quality-control evidence for a defined population, but different anchor types, substrates, installers or site conditions may require separate consideration.
Primary Technical References
The main technical references for this subject are BSI information relating to BS 8539 and the Construction Fixings Association guidance notes. Manufacturer-specific installation and technical documentation should also be checked for the anchor system being installed. The detailed provisions of BS 8539 should be checked in the official licensed British Standard. This article does not reproduce the complete anchor selection, design, installation or testing procedure contained in the standard.
Source Context and Technical Note
This article is a STRUCTinspect technical explainer covering practical considerations when specifying and procuring post-installed anchor testing in UK construction. Anchor testing should not be used as a substitute for appropriate anchor selection, structural design, competent installation, inspection or manufacturer requirements. A site test can only answer the engineering question for which it was properly specified and carried out.
This article does not provide structural design, temporary works, legal or contractual advice. Anchor selection, test loads, sampling regimes, acceptance criteria and remedial actions should be confirmed by the engineer, designer, specification author or other competent professional responsible for the relevant fixing system.
