Bridge & Bedrock

Every earthworks specification, however carefully written, is only a promise on paper until it is checked against what actually happens on site. Construction Quality Assurance, universally known by its acronym CQA, is the structured programme of inspection, sampling, testing, and documentation that verifies a specification has genuinely been met. It sits at the intersection of geotechnical engineering, contract administration, and site supervision, and it is often the least glamorous part of a project — yet it is also the part that determines whether a scheme performs as designed for its entire service life. This post looks at why CQA is needed, how testing frequency is decided, what a validation report actually contains, and the common failures that undermine CQA programmes in practice.

Why It’s Needed

A geotechnical design is built on assumptions: assumed material properties, assumed compaction standards, assumed groundwater conditions. None of these assumptions are self-fulfilling. Without a formal process for checking that what is actually placed and built on site matches what the designer assumed, there is no way of knowing whether the finished structure will behave as intended. CQA exists to close this gap between design intent and constructed reality, converting an assumption into a verified fact, backed by evidence that can be produced years later if a dispute or a performance question ever arises.

The consequences of skipping this step are not abstract. Under-compacted fill, moisture contents outside the acceptable range, or material that does not meet the specified classification can all be placed and buried without any visible sign of a problem at the time. The defect only becomes apparent later, sometimes years later, as settlement, cracking, instability, or serviceability failure. By then the offending material is deep within a completed structure, and remediation is vastly more expensive and disruptive than it would have been if the problem had been caught during construction. CQA is, in this sense, a form of risk management: it trades a modest, ongoing cost during construction for a large reduction in the probability and cost of failure later.

Independence matters enormously here. Where the party responsible for testing compliance is the same party responsible for building the works, there is an inherent conflict of interest, however well-intentioned everyone involved might be. Good CQA regimes are structured so that testing and verification are performed or overseen by a party independent of the construction team, reporting to the client, the engineer, or in regulated sectors such as landfill engineering, to the environmental regulator. This independence is what gives a CQA record its credibility: it is evidence that the works were checked by someone with no incentive to overlook a failing result.

Testing Frequency

How often should a layer of fill be tested? Too infrequent, and a localised area of poor compaction can slip through undetected; too frequent, and the cost and programme impact of testing becomes disproportionate to the risk being managed. UK earthworks specifications typically address this with minimum testing frequency tables, expressed as one test per given plan area or per given volume of material placed, with the frequency varying by material type and by the consequence of non-compliance. Structural fill beneath a foundation, for example, will generally be tested far more frequently than general landscaping fill in a low-risk area, reflecting the very different cost of getting it wrong.

Frequency is not applied uniformly through a project either. Early in the works, when a contractor’s plant, methods, and workforce are unproven for the particular material being handled, testing is often intensified to build confidence that the chosen method is capable of achieving the specification consistently. Once a track record of compliant results has been established, and provided nothing changes in the source material, plant, or personnel, frequency may be relaxed towards the baseline set out in the specification. Any non-conforming result should trigger the opposite response: increased frequency in the affected area, investigation of the cause, and a hold point preventing further work from being covered up until the issue is resolved.

Different tests carry different frequency requirements because they measure different things at different speeds. Field density and moisture content tests, using methods such as nuclear density gauges or sand replacement, are fast enough to be carried out on a large proportion of layers and are often the workhorse of routine compaction verification. Plate bearing tests and in-situ CBR testing, which take longer and disturb a larger area, are typically used more sparingly, often at specific hold points or on a sampling basis. Laboratory index tests such as particle size distribution and Atterberg limits are slower still, and are generally used to confirm that a material remains within its classified type rather than as a routine layer-by-layer check.

Validation Reports

The output of a CQA programme is not just a pass or fail on the day a test is carried out; it is a permanent, auditable record known as the validation report. This document draws together every test result, every non-conformance and how it was resolved, calibration certificates for the equipment used, as-built records showing what was actually constructed and where, and a formal statement of compliance from the engineer responsible for the CQA regime. In regulated sectors such as landfill engineering, a validation report is often a legal requirement before an environmental permit will allow waste to be accepted into a newly constructed cell, and the report must satisfy a regulator, not just a client.

A well-structured validation report typically follows a consistent format: an introduction setting out the scope and specification against which the works were assessed, a methodology section describing how testing was carried out and by whom, results presented systematically by area or chainage so that coverage can be checked at a glance, a clear account of any exceptions or non-conformances and the remedial action taken, and a set of appendices containing the raw data behind every summary table and chart. This structure matters because a validation report is written for readers who were not present during construction: future asset owners, insurers, regulators, and engineers designing later phases or additions who need to understand the ground conditions that were actually achieved, not merely those that were intended.

The value of a good validation report often only becomes apparent long after construction has finished, when a dispute arises, when a new structure is proposed nearby, or when an unexpected settlement prompts an investigation into whether the original earthworks were adequately controlled. A thorough, well-organised report can resolve such questions quickly and conclusively; a poor one, or worse, no report at all, leaves everyone guessing about ground conditions that are now buried and unverifiable.

Common Failures

CQA programmes fail in fairly predictable ways, and recognising them is the first step to avoiding them. The most fundamental is a loss of independence, where testing personnel are placed under commercial or programme pressure by the party whose work they are checking, whether through direct instruction or simply through the discomfort of repeatedly reporting bad news to the people paying their invoices. Closely related is the practice, sometimes unconscious, of selecting test locations for convenience rather than at random or in accordance with a proper sampling plan, which can turn a testing regime into a demonstration exercise rather than a genuine check.

Documentation failures are just as damaging as testing failures, even though they attract less attention at the time. Non-conformances that are identified but never formally closed out, calibration certificates that are allowed to lapse, records that are incomplete or scattered across different systems, and a lack of continuity when site personnel change all erode the reliability of the final record. A CQA regime that looks rigorous in the field can still produce a validation report that is effectively worthless if the paperwork behind it cannot be trusted or reconstructed.

Programme pressure towards the end of a job is another recurring theme. As completion dates approach, there is a natural temptation to relax hold points, accept marginal results without full investigation, or backfill testing gaps retrospectively rather than delaying the works. This is precisely when discipline matters most, because the areas completed under time pressure are often the last to be checked and the first to be forgotten once the site has been handed over. A CQA programme is only as strong as its weakest moment, and that moment is usually the one when everyone is most eager to move on.

None of this is a case for bureaucracy for its own sake. A good CQA programme is proportionate, independent, and honest: it tests enough to genuinely verify compliance, it is free from commercial pressure to produce a particular answer, and it produces a record that can be trusted long after everyone involved in the original works has moved on. Earthworks specifications tell a contractor what to build; CQA is how everyone else finds out whether they actually did.