QL-A is expensive certainty, used sparingly
Physical exposure is the only way to turn an interpreted position into a measured one. PAS 128 calls that Type A and quality level QL-A. It is not a substitute for Type B detection and it is not something you do on every linear metre of a UK site. You do it where a clash, a cover level, a pile cap or a connection cannot be designed on a quality-levelled guess.
A trial hole that is not tied back to the survey grid is just a hole. The value is the measurement: coordinates, level, diameter, material, and a photograph that will still make sense in a year. That information is written onto the same utility mapping segments the designer already holds, so QL-A is an upgrade to a line, not a separate PDF.
HSG47 remains the safety document for the act of exposing the service. PAS 128 is the language for reporting what you found. Confusing the two is how teams either over-dig or under-record.
Choosing which crossings deserve a hole
Start with the Type D plot and the Type B drawing. Look for crossings that sit inside a proposed excavation, services with conflicting records, and apparatus whose depth would change a design if it were half a metre out. Those are QL-A candidates. A hole opened because “we always pothole everything” is a budget, not a specification.
The Elizabeth line corridor and later over-site development in London are the usual illustration: a few interfaces really do need exposure; the rest of the block can stay at detection quality. Birmingham city-centre regeneration has the same pattern around podiums and highway closures — a small number of crossings control the programme.
What we need before we open the ground
Permissions, a traffic or possession plan if the hole is in a live surface, a service owner’s attendance rules, and a reinstatement specification. We will not invent a street works notice or a utility owner’s consent. If those are not in place, the survey stops at QL-B and the residual-risk note says so.
Read trial holes versus non-intrusive detection if you are still deciding how many holes a Leeds or Cardiff plot actually needs. Then send a drawing with the candidate crossings marked. That is a better brief than “full potholing”.
Reinstatement, records and the hole you do not open
A trial hole that is opened and then poorly reinstated becomes a new defect. The specification for backfill, surface and any highway authority rules should sit in the instruction, not in a conversation after the tarmac has been cut. We will not invent a street-works notice. If the notice is missing, the hole waits.
Photography should be boring and complete: context, the exposed apparatus, a scale, and the coordinated marks. Artistic site photography is how details disappear. These are not marketing images and we will not reuse them as hero shots claiming a portfolio we have not agreed to publish.
The hole you do not open is also a decision. If Type B already supports the design, spending money on exposure to “be sure” about every service is how budgets vanish. Write the decision down. Future you, standing in a progress meeting, will want to know why that crossing stayed at QL-B.
Elizabeth line corridor interfaces and Birmingham regeneration podiums were named because they concentrate expensive clashes in a small number of points. A campus spine in Leeds or a Bay plot in Cardiff can look the same once you mark the pile line. The geography changes. The logic of QL-A does not.
Type A still sits on top of Type D. Opening a hole without a records search is how teams expose the wrong owner’s plant and then discover the one they cared about a metre away. Sequence is cheaper than courage.
When the hole is closed, update the model the same week. A QL-A measurement that lives only in a WhatsApp photograph is not a PAS 128 upgrade. The segment on the coordinated drawing is the product you paid for.