PAS 128 Utility Surveys — Quality Levels QL-A to QL-D explained

What PAS 128 is, and why it exists

PAS 128 is the UK Publicly Available Specification for the detection, verification and location of buried utilities, published by BSI and updated in 2022. It was created because UK utility surveys had no common quality framework — clients could not compare bids, surveyors used inconsistent methods, and downstream design teams had no way of knowing how much trust to place in a given record. PAS 128 introduced a four-tier quality scale so that any buried-asset position can be described with a defined level of confidence, from a desktop records search at one end to a physically exposed and surveyed coordinate at the other.

The four Quality Levels

The table below summarises the four quality levels. The accuracy values are indicative — always check the specific tier and tolerance required by your client brief, as specifications vary.

How QL maps to CDM 2015 Regulation 11(6)

Regulation 11(6) of the Construction (Design and Management) Regulations 2015 places a duty on the principal designer to identify, eliminate and control foreseeable risks — including the risk of striking buried services — so far as is reasonably practicable. The PAS 128 quality framework gives that duty a measurable anchor: the higher the quality level used for a given asset, the more defensible the claim that risk has been investigated and controlled.

In practice, the principal designer must decide on a per-asset basis what level of investigation is reasonably practicable. A QL-D records search is a reasonable starting point for a low-risk small-diameter pavement intervention. A QL-A verification is usually expected at the location of any planned deep excavation, any proposed pile, and any utility crossing that cannot be designed around. The proportionality test is recorded in the pre-construction information.

The same principle is reflected in our own work on NUAR and the CDM liability — NUAR is QL-D data, and treating it as anything more without explicit verification is one of the most common compliance gaps we encounter on audit. The companion guide to CDM 2015 Regulation 11(6) compliance for buried services walks through that test in more detail.

Choosing the right QL for your project

The decision is rarely a single QL across an entire site. A phased strategy is usually more proportionate and more economic. The flow we recommend is:

1. Start with QL-D. Pull statutory undertaker records via NUAR for the full site boundary. Identify all known utility runs, ownership, and any recorded depths. Treat depth and abandoned-asset records with caution.
2. Validate at QL-C. Walkover the site to correlate records with visible covers, valves, marker posts and surface scars. Identify records that are clearly out of date.
3. Detect at QL-B. Commission a geophysical survey of the works footprint. Specify the QL-B tier required for the level of detail your design needs.
4. Verify at QL-A where it matters. Use vacuum excavation or trial pits at planned excavation locations, crossings of high-consequence assets, and any position where a small error would cause a strike.

Common mis-assumptions and their consequences

“NUAR records are good enough for design.”

NUAR aggregates records from over 700 asset owners but does not improve the underlying record quality. In PAS 128 terms it is QL-D. Designing depth-sensitive works against QL-D records, without QL-B detection at the works footprint, is a recurring cause of strike incidents.

“A QL-B survey covers the whole site.”

QL-B detection is constrained by the equipment, the surface conditions and the material of the asset. Non-metallic pipes, deep services, and assets below reinforced concrete may be invisible to the survey. The survey report should list what was not detectable, not just what was detected.

“Abandoned assets do not matter.”

Many recorded abandoned assets remain pressurised, live, or contaminated. Treating “abandoned” as a reason to discount the asset from risk assessment is a frequent enforcement theme. The default position is that the abandonment status itself is a quality-of-record question until verified.

“Depth from QL-B is reliable.”

Depth estimates from electromagnetic locators degrade rapidly with depth and with congestion of other services. A QL-B depth is best treated as indicative; designs that depend on a precise depth should specify QL-A verification at the design point.

How NHDM uses QL as a data-quality modifier

The National Hidden Data Map (NHDM) is a buried-services risk scoring tool built by ActiveSense. It produces an asset-level risk score for any polygon the user defines, using the formula consequence × proximity × data_quality_modifier. The data-quality term is anchored on the PAS 128 quality level of each asset record: QL-A is 1.0 (no uplift), QL-B is 1.2, QL-C is 1.4, QL-D is 1.7, and Unknown is 2.0. Records with no logged QL are treated as Unknown, which is usually a worse score than a deliberate QL-D.

The effect is that two sites with identical asset density and identical consequence profiles can produce very different overall risk scores if one site relies on QL-D records and the other has been verified to QL-A at known crossings. The output is a CDM-aligned PDF report (Regulation 11(6) framing) that flags every asset in the works footprint, the QL it relies on, and the proximity zone in which it sits. It is one of the capabilities we are developing for utility-strike-risk projects.

FAQ

Is PAS 128 a legal requirement in the UK?

PAS 128 is a publicly available specification published by BSI, not a statute. It is, however, widely written into client briefs, framework agreements and Highways England / National Highways requirements, and is referenced as good practice in HSE guidance. For projects subject to CDM 2015, a PAS 128 survey is one of the most defensible ways of demonstrating that buried-services risk has been identified so far as is reasonably practicable.

What is the difference between PAS 128 QL-B and QL-A?

QL-B is a detection survey, typically using electromagnetic locators and ground-penetrating radar to identify and trace a utility from the surface. QL-A is verification of position by physical exposure, typically by vacuum excavation or trial pit, with horizontal and vertical accuracy generally specified as ±15 mm in many client specifications. QL-A confirms what QL-B detected.

When is QL-D survey appropriate on its own?

QL-D, which is a desktop records search of utility owner data, is appropriate as a preliminary screening exercise, for low-risk feasibility work, or as the first stage in a phased survey strategy. It is rarely appropriate on its own for construction works because record accuracy is highly variable, depths are usually absent, and abandoned assets are often unrecorded. Using QL-D alone for buried-services risk on a CDM project is one of the most common audit findings.

How does NUAR relate to PAS 128?

The National Underground Asset Register (NUAR) aggregates utility records from over 700 asset owners into a single platform. In PAS 128 terms, NUAR is a QL-D data source. It does not replace QL-C, QL-B or QL-A site survey work. NUAR is a faster and more complete way of obtaining records than individual statutory undertaker requests, but the records themselves carry the same data-quality limitations.

What accuracy can I expect from QL-B?

PAS 128 specifies tiered horizontal accuracy bands for QL-B, with QL-B1 being the most accurate detection result and QL-B4 the least. Typical horizontal accuracies referenced in client specifications range from approximately ±150 mm to ±500 mm depending on the band, with depth accuracy generally less reliable than plan position. Always check the specific tier required in your brief.

Does a PAS 128 survey discharge the principal designer CDM duty?

No. A PAS 128 survey is evidence that buried-services risk has been investigated, but the principal designer remains responsible for ensuring that the information is integrated into pre-construction information, that residual risks are communicated to the principal contractor, and that the survey scope was appropriate to the risk profile. The survey is an input to the duty, not a discharge of it.

How should QL be recorded in GIS or data systems?

Each asset record should carry its source quality level as an attribute (typically QL-A, QL-B, QL-C, QL-D or Unknown), the date of survey, the survey company, and the method used. This metadata is essential for downstream risk scoring, for CDM handover packs, and for any subsequent design or excavation decision. Without it, every asset is effectively Unknown quality.