EML follows current. Live cables and conductive pipes give it something to lock onto. Plastic water, many fibre ducts and abandoned non-metallic features do not. GPR follows contrast. It can image those quiet assets, and it can also show voids and changes in made ground. It does not attribute a live LV cable with the same confidence a locator can. That is the whole pairing.
PAS 128:2022 writes the pairing into the Type B minimum, alongside Type D and Type C foundations. The CICES explanation repeats it in public language. If a quotation offers “PAS 128 Type B” and lists only a cable-avoidance tool, the quotation is describing a different product.
There is a legitimate radar-only task: increasing confidence in Type D information so a later Type B can be targeted. That is a scoping tool. It is not a Type B survey. Likewise, a CAT and genny sweep before excavation is an HSG47 duty, not a substitute for a quality-levelled Type B model.
Ground conditions change the balance, not the requirement. Wet clay shortens radar depth. Congested streets confuse locators. The professional response is a residual-risk note and, where justified, a trial hole — not the quiet deletion of a method to protect a price.
If you are writing a specification this week, name both methods, name the methodology type, and say whether GPR will be post-processed. Then read the service pages for GPR surveys and electromagnetic locating. The PAS 128 page shows how they sit in the full sequence.
Budgets try to split the pair because locators are familiar and radar looks expensive. The expensive event is the plastic water main nobody traced. Saving the radar fee to pay for that repair is not a saving. Write the pair as one line item called Type B so nobody can delete half of it in a value-engineering meeting.
There are plots where one method will outperform the other. A dry sandy verge may give radar a good day. A congested live street may give EML a noisy one. The specification still wants both, because the poor day is the one that hides the service you care about.
Array systems and processed cubes are tools, not a different specification. They can raise confidence and still leave a quality level that is not QL-A. Do not let a demonstration video replace a method statement.
If a supplier offers “PAS 128 Type B (EML only)” as a named product, ask them to show where the 2022 text allows that. The public CICES note is a reasonable place to start the conversation before anyone mobilises.
Training days that only teach cable-avoidance tools are still useful for excavation. They are not a Type B survey course. Keep the two conversations on two slides so a site manager does not think a CAT ticket is a PAS 128 appointment.
When ground defeats one method, the other is why the pair exists. Write the poor result down. Deleting the poor method from the next quotation is how the next project repeats the miss.
If you need a shorter internal note, send this page and the GPR and EML service pages together. Three links are cheaper than a misunderstood one-line specification.
Type B is a pair because the ground is mixed. The UK’s buried estate is mixed. A specification that picks a favourite method is specifying a favourite blindness. Keep both methods. Keep the residual-risk note. Then excavate as HSG47 still requires.
Take this into the brief
Keep a short internal rule: if a quotation deletes either electromagnetic locating or ground penetrating radar from a Type B line item, it is not a 2022 Type B quotation. Ask for the pair to be restored or for the product to be renamed. Names matter because later people will copy them into a construction-phase plan.
On site, the pair also protects the surveyor’s own honesty. A locator-only day on a street full of plastic water produces a tidy drawing of the metallic network and a silence where the dangerous miss will sit. A radar-only day on a street full of live cables produces shapes without attribution. Writing both results onto one model is how UK teams stop choosing a favourite blindness.