A chemical anchor, also known as a bonded or resin anchor, fixes a threaded rod, reinforcing bar or internally threaded socket into a drilled hole in concrete or masonry using a two-part reactive resin. The resin bonds to both the steel and the wall of the hole, so load is carried along the full embedment length rather than by an expansion sleeve pressing outwards against the base material.
That bond is why chemical anchors are specified close to edges, in tight groups, in hollow and weak masonry and for post-installed rebar, and it is also why they are less forgiving on site than mechanical fixings. In the UK, selection, installation and testing are covered by BS 8539:2012+A1:2021, design of fastenings in concrete by BS EN 1992-4:2018 and its UK National Annex, and the performance of each individual product by its European Technical Assessment (ETA).
Cross-section illustrating a threaded rod bonded into concrete using chemical anchor resin. Illustration: London Construction Magazine.
Key Takeaway: "Chemical anchor" is not a performance class. Cracked concrete, sustained tension, fire, seismic loading and wet holes are each acceptable only where the specific product's ETA covers them, and the bond that ETA promises is only achieved if the hole is drilled, cleaned and filled exactly as the manufacturer instructs.
How Bond Differs From Expansion
A mechanical expansion anchor works by forcing a wedge or sleeve against the borehole wall, generating friction and local bearing stress in the concrete. Those radial forces are what limit minimum edge distances and spacings, because they can split the base material when an anchor is placed too close to a free edge or to its neighbours.
A bonded anchor generates no expansion force at all. Load passes through adhesion and mechanical keying between the cured resin and the roughness of the drilled hole, which distributes stress along the embedded length and allows deeper or variable embedment within the limits of the ETA. The trade-off is that capacity depends on things a mechanical anchor is less sensitive to: dust left in the hole, how well the two components have mixed, the temperature of the base material and whether the anchor was disturbed before the resin cured.
Resin Types and Where Each Is Used
Most site work uses injection mortars dispensed from two-part cartridges through a static mixing nozzle. The chemistry names describe broad families rather than guaranteed performance, so the table below is orientation only and the ETA for the actual product always governs.
| System | Typical use | Cure behaviour | What to check in the ETA |
|---|---|---|---|
| Polyester | Lighter-duty and non-structural fixings in masonry and non-cracked concrete | Fast | Often limited to non-cracked concrete; generally not the choice for sustained tension or permanently damp conditions |
| Vinylester (epoxy acrylate) | General structural fixings, M&E supports, façade and plant brackets | Faster than epoxy | Cracked-concrete, damp-hole, seismic and fire coverage vary between products |
| Hybrid | Structural and M&E work where quicker loading is needed | Rapid strength gain, still temperature-dependent | A manufacturer descriptor rather than one chemistry; long-term load performance must be confirmed |
| Epoxy | Heavy structural connections, post-installed rebar, deep embedment, sustained tension | Slowest; long working time | Sustained-load suitability and minimum installation temperature; fast-set variants are not interchangeable with standard epoxies |
| Capsule | Repetitive fixed-dose fixings into solid concrete or masonry | Mixed in the hole as the stud is driven | Hole diameter and depth must match the capsule; setting tool, rotation and drive time are product-specific |
Hollow brick, perforated brick and hollow blockwork need a different approach because injected resin simply runs into the voids. A perforated plastic or steel mesh sleeve is inserted first, and the resin is injected into it so that it squeezes through the perforations and forms a keyed plug behind the face shell. Capacity then depends on the sleeve, resin and rod being used as the tested combination, so swapping any one component can invalidate the assessed system.
What the ETA Does and Does Not Cover
An ETA is issued by a technical assessment body against a European Assessment Document; for bonded anchors in concrete this is EOTA's EAD 330499 for bonded fasteners. The ETA declares assessed performance for defined conditions, including tension and shear resistance, cracked or non-cracked concrete, embedment range, permitted drilling methods, temperature ranges, installation parameters and, where assessed, fire and seismic performance.
What it does not do is confirm that a product suits a particular structure, substrate condition or load combination. That remains a design decision under BS 8539, and CE or UKCA marking only confirms that the manufacturer declares conformity with the assessed performance. If a condition is not stated in the ETA or the manufacturer's data, the product should be treated as unqualified for it until further assessment or site testing says otherwise. LCM's guide to the role of ETAs, EADs and CE/UKCA marking in anchor design covers the approval chain in more detail.
Installation Sequence
BS 8539 expects installers to be trained in the specific anchor type, supervised, and to follow the manufacturer's installation instructions in full. The exact sequence differs between products, but a typical injection installation runs as follows.
- Check the specification against what is on site: product, ETA reference, rod diameter, embedment depth, edge and spacing distances, substrate and the cracked or non-cracked condition.
- Drill to the specified diameter and depth using a drilling method the ETA permits. Some products are assessed for hammer drilling only, others for diamond coring or hollow drill bits with dust extraction. If reinforcement is struck, stop and refer back to the designer rather than relocating the hole unilaterally.
- Clean the hole using the manufacturer's blow–brush–blow sequence: blow out from the bottom of the hole, brush with a correctly sized steel brush that still resists insertion, then blow again. The number of cycles is set by the instructions and ETA, with larger diameters or deeper holes sometimes requiring more.
- Fit a new static mixer and discard the first resin until it extrudes in a uniform colour, showing that the components are fully mixed.
- Inject from the bottom of the hole, withdrawing the nozzle slowly to avoid trapping air, to the fill level stated in the instructions, using an extension tube for deep holes.
- Insert the clean, oil-free rod or rebar with a slight twisting action to the marked embedment depth, checking that resin appears at the collar.
- Leave the anchor undisturbed through the gel time, and only fit the fixture and apply the specified torque once the full cure time has elapsed.
Damp, wet or water-filled holes are acceptable only where the ETA explicitly covers that condition, and those products often carry a separate cleaning and injection procedure. Where dust-extraction drilling is used and the instructions state that no further cleaning is required, that instruction applies only to the product and drill system it was written for. LCM has looked separately at why resin anchors fail when hole cleaning is ignored.
Gel Time, Cure Time and Temperature
Gel time is the working window in which the rod can be inserted and aligned; full cure time is the point at which the anchor can be torqued and loaded. Both depend on the temperature of the base material rather than the air, and the spread between products is wide. One published ETA for an injection mortar gives a minimum curing time of 17 hours at between -20°C and -16°C, 50 minutes at 0°C to +4°C, 15 minutes at +10°C to +19°C and six minutes at +30°C to +40°C, while other products cannot be installed at all below +5°C.
Those figures show the scale of variation, not values to plan around. On a British winter site, the difference between a product's 20°C cure time and its cure time in a cold slab can run to many hours, and the cartridge itself also has a permitted storage and use temperature range. Site teams need the current temperature table for the exact cartridge being used, and a way of knowing the substrate temperature, before programming the loading of anchors.
Installation Errors Behind Most Failures
Construction Fixings Association guidance identifies bonded anchors as particularly sensitive to dust left in the hole, because the resin bonds to a loose dust layer rather than to sound concrete. Incomplete cleaning, or a worn or undersized brush that no longer scrapes the wall, is the error most often linked to underperforming resin anchors.
The other recurring causes are just as basic. Unmixed resin left in the nozzle can leave soft material in the hole; injecting from the top traps air and reduces the bonded area; moving the rod after gel time breaks the developing bond; and loading or torquing before full cure, or over-torquing with an impact driver, damages the adhesive layer. Expired cartridges, reduced embedment, oversized holes from worn bits and the use of a product outside its ETA scope for cracked concrete, flooded holes or sustained tension complete the list. Most of these are invisible once the fixture is fitted, which is why BS 8539 places so much weight on supervision and on the supervisor signing off each installation before it is loaded.
Site Testing Under BS 8539
BS 8539 separates tests that establish an allowable resistance from tests that check installation quality. Preliminary tests are used where no ETA covers the anchor and base material combination, or where the substrate falls outside the ETA's scope, and they are carried out on dedicated test anchors to derive a design resistance statistically. Proof tests are applied to a sample of the working anchors to confirm workmanship, and they do not create a design resistance.
For proof testing, the regime commonly applied under BS 8539 tests at least 2.5% of installed anchors, with a minimum of three tests in each discrete area, at a load of 1.5 times the characteristic action on the anchor. Testing 5% allows the factor to drop to 1.25, and the proof-load factor should not exceed 1.5. A single failure in an area means investigating the cause and doubling the sample to 5%, with at least six tests; more than one failure means testing every anchor in that area and reconsidering the specification. A discrete area is defined by anchor type, base material or a change in conditions such as flooded holes, so a single project can need several separate samples.
These are the figures most specifications follow, but the governing regime is whatever the designer or specifier sets for the project, and the test load must never exceed the manufacturer's recommended resistance. Testing should be carried out by a competent tester with calibrated equipment, following the CFA's guidance note on site testing construction fixings, and each result needs recording against the anchor location, product, ETA reference, substrate condition, installer and tester. LCM's BS 8539 anchor testing and compliance guidance sets out the full compliance framework, including the evidence needed on higher-risk buildings, and the distinction between test types is covered in pull-out testing vs proof testing under BS 8539 and BS 7883.
Fall Protection Anchors and BS 7883
Where an anchor forms part of a personal fall protection system, BS 7883:2019 applies to the system as a whole. It requires the system designer to confirm that the supporting structure is adequate, calls for trial tests of structural anchors in the same base material, with the calculated test load held for at least three minutes without failure, and requires a System Technical File and an inspection scheme for the life of the system. BS 8539 proof-test sample rates should therefore not be applied to fall protection anchors without checking the BS 7883 requirements and the system designer's instructions.
Lessons From the Boston Tunnel Collapse
The most thoroughly investigated bonded-anchor failure remains the collapse of suspended concrete ceiling panels in Boston's Interstate 90 connector tunnel on 10 July 2006, which killed a passenger in a car below. The US National Transportation Safety Board concluded that the epoxy used to bond the ceiling-support anchors had poor creep resistance, so it deformed progressively under sustained tension until anchors pulled free. According to the NTSB's safety recommendation letter, Federal Highway Administration testing found continuing displacement at loads as low as 1,000 lb against a design service load of 2,600 lb, and the board found the fast-set epoxy unsuitable for long-term tension.
The adhesive had not failed a short-term test; it failed a load case it was never suitable for. The design process did not identify sustained tension and creep as the governing condition, which is the same selection step BS 8539 places with the designer and specifier today. For UK projects, the practical lesson is that the ETA must explicitly cover sustained load for any permanently tensioned bonded anchor, particularly overhead, and that a contractor substituting a faster or cheaper resin without specifier approval takes on the design responsibility for that anchor.
Those responsibilities, and the records that support them, also run into CDM 2015 and, on higher-risk buildings, the Building Safety Act golden thread; LCM has set out which anchor records must be kept under the Building Safety Act. The code itself is due for change: the BSI committee has proposed revising BS 8539 because the documents it relies on for site testing, European Technical Report TR053 and the CFA site testing guidance note, have since been amended, including new guidance on testing in masonry.
Frequently Asked Questions
Can chemical anchors be installed in wet or flooded holes?
Only where the product's ETA explicitly covers damp or water-filled holes, and then only using the manufacturer's cleaning and injection procedure for that condition. Products without that coverage should not be used in wet holes.
Only where the product's ETA explicitly covers damp or water-filled holes, and then only using the manufacturer's cleaning and injection procedure for that condition. Products without that coverage should not be used in wet holes.
Is site testing always required under BS 8539?
No. Preliminary tests are needed where no ETA covers the anchor in the actual base material, and proof tests where the specification requires them, which is typical for safety-critical anchors. BS 8539 is a code of practice rather than legislation, but it is the recognised standard of care for safety-critical post-installed anchors in the UK.
No. Preliminary tests are needed where no ETA covers the anchor in the actual base material, and proof tests where the specification requires them, which is typical for safety-critical anchors. BS 8539 is a code of practice rather than legislation, but it is the recognised standard of care for safety-critical post-installed anchors in the UK.
How long before a chemical anchor can be loaded?
Not until the full cure time for that product at the actual base-material temperature has passed. Depending on the resin and the temperature, that can range from minutes to more than a day, so it must be read from the current manufacturer table for the cartridge in use.
Not until the full cure time for that product at the actual base-material temperature has passed. Depending on the resin and the temperature, that can range from minutes to more than a day, so it must be read from the current manufacturer table for the cartridge in use.
Expert Verification & Authorship
Founder & Author, London Construction Magazine | Construction Testing & Investigation Specialist