Proof Load Testing vs Ultimate Load Testing: When to Use Each Method

Every anchor installation carries two fundamental questions: has this anchor been installed correctly, and how much load can it actually carry before it fails? These questions sound similar but they require different test methods to answer. Proof load testing addresses the first question by applying a defined load below the anchor's design capacity and confirming the anchor holds without measurable movement. Ultimate load testing addresses the second by loading the anchor to destruction and recording the failure load, failure mode, and displacement behaviour throughout. Choosing the wrong method for the wrong purpose wastes money at best and leaves genuine safety questions unanswered at worst.

In Australian practice, both methods are governed by different clauses within the same standards framework. AS 5532:2013 (currently under revision, with AS 5532:2025 expected to supersede it) covers the installation and testing of anchors used in height safety systems, while AS/NZS 1891.4:2009 (updated as AS/NZS 1891.4:2025) addresses the selection, use, and maintenance of industrial fall arrest systems including their anchors. The AEFAC Technical Note TN05 provides additional guidance on anchor testing protocols for engineered systems. Understanding where each test method fits within this framework is what determines whether a test programme actually proves what the certifying engineer needs it to prove.

This distinction matters practically because the two methods produce fundamentally different information. A proof test tells you that a specific installed anchor survived a specific load without pulling out or cracking its substrate. An ultimate test tells you the maximum load the anchor system can resist and how it behaves as it approaches and exceeds that limit. Neither replaces the other.

What Proof Load Testing Actually Does

Proof load testing applies a predetermined test load to an installed anchor, holds it for a specified duration, and measures displacement before, during, and after the load is applied. If the anchor holds the load within defined displacement tolerances and returns to near its original position after the load is released, the test is considered passed. No anchor is destroyed. The installation remains in service.

The test load in a proof load programme is typically set at a multiple of the working load limit (WLL). For fall arrest anchors rated at 6 kN under AS/NZS 1891.4:2025, proof test loads are commonly applied at 12 kN, representing a 2:1 factor. For anchors with higher rated capacities, such as 15 kN or 21 kN systems used in horizontal lifeline configurations, proof test loads scale accordingly. The exact load requirements depend on the standard being applied, the system configuration, and whether the anchor is being tested as a single point or as part of a loaded system.

Displacement is the critical measurement in proof testing. The anchor must not displace more than a defined tolerance during loading, and residual displacement after load removal must fall within acceptable limits. These values are typically measured in millimetres and recorded against the applied load at defined time intervals. An anchor that displaces 3 mm under proof load and does not recover more than 0.5 mm after load release is behaving differently from one that displaces 0.3 mm and fully recovers, and that difference tells the testing engineer something meaningful about the substrate's condition and the anchor's installation quality.

What Ultimate Load Testing Actually Does

Ultimate load testing loads an anchor incrementally until it fails. The failure might be anchor pullout, concrete cone failure, substrate splitting, anchor bar fracture, sleeve collapse, or chemical capsule bond failure, depending on the anchor type and substrate condition. The test records the full load-displacement curve from zero to failure, and the failure load itself becomes the primary output.

This is destructive testing by definition. The anchor tested cannot be returned to service. The surrounding substrate may be damaged. In reinforced concrete, a concrete cone failure at a M16 chemical anchor might extract a cone of concrete weighing several kilograms and leave a void 200-300 mm in diameter at the surface. In precast panels or hollowcore planks, the damage can extend further because the geometry of these elements concentrates stress differently than solid poured concrete.

Ultimate testing is typically performed during a design verification phase rather than as a quality assurance check on installed production anchors. An engineer specifying a new anchor system for a roof might commission ultimate load tests on trial anchors in representative substrate samples before finalising the design. This establishes the actual capacity envelope of that anchor in that specific concrete mix, age, and reinforcement configuration, rather than relying on manufacturer data derived from testing in idealised laboratory conditions.

When to Use Proof Load Testing

Proof load testing is the standard method for quality assurance on installed fall arrest and height safety anchors across Australian projects. It is non-destructive, which means every anchor in a system can be tested individually without sacrificing any of them. This is why AS/NZS 1891.4:2025 and AS 5532:2025 both contemplate proof testing as the default verification method for installed anchor systems.

The practical triggers for proof load testing include:

• New installations: : Every anchor in a new fall arrest system should be proof tested before the system is commissioned. This confirms that the installer has achieved correct embedment depth, torque, and substrate engagement across every point, not just the ones they were watched on.
• Periodic recertification: : AS/NZS 1891.4:2025 requires periodic inspection and testing of height safety systems at intervals not exceeding 12 months for most applications. Many anchor systems are proof tested at each recertification to confirm ongoing integrity.
• Post-incident verification: : After any event that may have loaded a fall arrest system, including a fall arrest itself or near-miss loading, the anchors involved should be proof tested before being returned to service.
• Substrate uncertainty: : When there is doubt about concrete strength, curing conditions, or reinforcement proximity, proof testing each anchor provides direct evidence of performance in the actual installed condition.
• Handover documentation: : Building owners and strata committees increasingly require proof test certificates as part of project handover documentation, giving the person conducting a business or undertaking (PCBU) a defensible baseline for their WHS obligations.

The load is applied using calibrated hydraulic equipment, typically a hollow-core hydraulic ram seated against a bearing plate, with the load transmitted through a test rod threaded into the anchor or clamped to the anchor's attachment point. All equipment should be calibrated to a traceable standard, with calibration records available for audit. The test report should record the anchor identifier, applied load, hold duration, displacement readings, and the engineer's assessment of pass or fail against the applicable criteria.

When to Use Ultimate Load Testing

Ultimate load testing belongs at the design verification and research end of the testing spectrum. It answers questions that proof testing cannot, because it takes the anchor through its full load-displacement response and into failure. This produces data that is essential in specific circumstances:

• Anchor design verification in novel substrates: : When an anchor is being installed in a substrate outside the manufacturer's tested range, such as a lightweight concrete mix, an aged masonry wall, or a precast element with unusual geometry, ultimate testing in representative samples establishes actual capacity.
• Investigating existing anchor systems with unknown installation history: : If a building's anchor system has no documentation and the substrate is uncertain, ultimate testing of a sample of anchors can establish whether the as-installed capacity is adequate, although this approach requires careful statistical sampling because it destroys the tested anchors.
• Calibrating reduction factors for engineering assessments: : When a structural engineer is preparing a formal anchor capacity assessment under the National Construction Code (NCC) framework, they may need test-derived data to support the reduction factors applied in their calculations.
• Failure mode investigation: : After a failure or near-failure event, ultimate testing of anchors removed from the same batch or same substrate area can identify whether the failure was an outlier or a systemic problem.
• Research and product development: : Anchor manufacturers and testing laboratories conduct ultimate load testing as part of product development and standards compliance testing. This is where the rated capacities that appear on product data sheets originate.

Defining the Test Scope for Ultimate Testing

Because ultimate testing destroys each anchor tested, the scope must be defined carefully. Testing too few anchors produces data that cannot be used statistically with confidence. Testing in substrate that does not represent the actual installation conditions produces data that does not apply to the real installation. A minimum of five tests per variable set is a common starting point in engineering practice, though project-specific requirements may demand more depending on the variability observed and the consequence of under-performance.

The test setup for ultimate loading mirrors the proof test setup in its hardware but differs in its protocol. The load is applied in controlled increments, with displacement recorded at each increment. The increment size is typically 5-10% of the expected failure load, and the load is applied slowly enough to allow the substrate to respond without dynamic amplification. The load-displacement curve is plotted in real time, and the test continues until either the anchor fails completely or the test rig reaches its capacity.

How the Two Methods Compare Directly

It helps to set the two methods side by side across the criteria that matter most in project planning:

• Anchor fate: : Proof testing is non-destructive; ultimate testing destroys the anchor.
• Output: : Proof testing confirms pass/fail against a defined load; ultimate testing produces a complete load-displacement curve and a failure load.
• Application: : Proof testing is used for installation QA, recertification, and compliance verification; ultimate testing is used for design verification, capacity research, and failure investigation.
• Cost per anchor: : Proof testing is lower cost per anchor because no replacement is required; ultimate testing requires anchor replacement and may require substrate repair.
• Sample size: : Proof testing can and should cover 100% of installed anchors; ultimate testing covers a small sample by necessity.
• Standards reference: : Proof testing procedures are detailed in AS 5532:2025 and AS/NZS 1891.4:2025; ultimate testing protocols draw on BS 8539, AEFAC TN05, and project-specific test plans.
• Engineer involvement: : Both methods require involvement from a competent person, but ultimate test interpretation typically requires a structural or geotechnical engineer given the failure mode analysis involved.

Reading Displacement Data in Context

Both test methods produce displacement data, but the interpretation differs. In proof testing, displacement tolerances are pass/fail thresholds. In ultimate testing, displacement data tells the story of how the anchor fails.

An anchor that shows linear elastic displacement up to a high load and then fails suddenly through anchor bar fracture is behaving differently from one that shows progressive nonlinear displacement at moderate load followed by gradual pullout. The first failure mode suggests the anchor and substrate are well-matched, with the anchor itself being the limiting element. The second suggests substrate distress, possibly because of a low-strength concrete mix, inadequate embedment depth, or proximity to an edge or joint.

In hollowcore plank substrates, displacement behaviour is particularly important to monitor because the geometry of the voids creates stress concentration effects that differ from solid concrete. An anchor installed into the web of a hollowcore plank behaves differently from one installed into the flange, and ultimate testing in representative sections of actual hollowcore product is the only reliable way to establish that difference quantitatively.

Calibration and Chain of Evidence

Neither test method is worth anything without proper calibration and documentation. The hydraulic ram, pressure gauge or load cell, and displacement gauges used in anchor testing must be calibrated to traceable standards, with certificates current at the time of testing. In Australian practice, NATA accreditation of the testing laboratory or testing organisation provides the highest level of assurance, though project-specific calibration records with documented traceability are the minimum acceptable standard.

The test report forms the chain of evidence between the physical test and the engineer's certification. It should identify every anchor by location or identifier, record the test equipment used and its calibration status, present the raw load and displacement data, state the acceptance criteria, and record the outcome for each anchor. A test report that cannot be traced back to a specific hydraulic rig calibrated on a specific date by a qualified technician is difficult to defend under WHS Regulation audit or in the event of an incident investigation.

Selecting the Right Method for Your Project

The decision between proof load testing and ultimate load testing is not usually a genuine choice for most building projects. For installed height safety anchors, proof testing is the correct method, and it is what the standards require. Ultimate testing becomes relevant when the design itself is being validated, when there is genuine uncertainty about substrate capacity that cannot be resolved by other means, or when a failure requires forensic investigation.

Building owners, facilities managers, and height safety installers should be asking whether their testing provider understands this distinction and applies it correctly. A test programme that applies proof loads to a sample of anchors while leaving the remainder untested does not give the PCBU confidence across the whole system. A programme that conducts ultimate testing on production anchors and destroys them in the process provides precise failure data at the cost of the system itself. Knowing which tool answers which question is what separates a credible anchor testing programme from one that generates paper without genuinely establishing safety.

For projects across New South Wales, Victoria, Queensland, and Western Australia, Anchor Testing Australia provides both proof load testing and ultimate load testing services calibrated to the current Australian Standards framework. If you are planning a new installation, scheduling recertification, or working through an anchor design verification, contact our team through [anchortesting.com.au/services/proof-load-testing](https://anchortesting.com.au/services/proof-load-testing) or [anchortesting.com.au/services/ultimate-load-testing](https://anchortesting.com.au/services/ultimate-load-testing) to discuss which method your project actually requires.