Skipping Type D is a false economy. A detection crew that arrives without a plotted record spends the first day confirming things a statutory owner already knew, and may still miss the owner who never replied. PAS 128:2022 treats the desktop study as a prerequisite, not a nice-to-have appendix.
Specify a desktop study when you are buying a site, when the last returns are older than your risk appetite, when private networks are likely, or when several designers will argue about the same street. Do not specify one as a substitute for location. HSG47 still expects you to locate on site before you excavate.
A usable QL-D product is a coordinated plot plus a list of outstanding owners. A zip of PDFs is not a product. If a client already holds robust Type D data, the surveyor can accept it and move on — the CICES note allows that — but “robust” is a professional judgement, not a hope.
Waterfronts and regeneration plots are the usual places where records are fragmented: Clyde-side Glasgow, Cardiff Bay, and London corridors with successive diversions. The study is editorial. It is still cheaper than a redesign.
Commission the desktop utility study as soon as you have a red line. Then decide, with evidence, whether Type B is justified on the whole plot or only on the working area. That is how you spend the next fee once.
Programmes that treat records as a same-week errand are the usual failure mode. Statutory owners do not work to your planning committee. Start the letters when you first draw the red line, not when the detection van is booked.
A study should also record silence. An unanswered owner is a residual risk. Plotting a blank as if the owner had confirmed “nothing here” is how later Type B work starts from a lie.
Client-supplied Type D data is allowed when it is robust. “We have a folder from 2019” is not automatically robust. Ask the surveyor to say yes or no, in writing, before you skip the fee.
When the plot exists, use it to decide where not to detect. A verge you will never excavate can stay at QL-D. The working area can take the Type B money. That is specification, not stinginess.
Keep a register of requests and returns. A study without a register becomes a zip of files nobody can defend. The register is the professional product as much as the plot.
Revisit the study when a new owner appears or when a diversion is known to have happened. Type D is not a one-time certificate. It is a snapshot with a date.
If the purchase dies at QL-D, file the plot anyway. The next person to look at that land will thank you, and you will have spent the smallest amount of money that could have killed a bad idea.
Desktop studies are unglamorous and they prevent the most expensive kind of detection day: the one that reconstructs a paper trail in a van. Specify them early. Plot the conflicts. Then decide, with a date on the file, whether anyone needs to leave the office.
Take this into the brief
Specify the desktop study when the red line first exists, not when the detection van is already booked. Statutory clocks do not care about your committee date. A register of requests and returns is part of the product. Silence from an owner is a residual risk, not a blank you may colour in.
If someone offers to “skip the paper and scan it”, ask them which owner they are willing to miss. Type D is unglamorous and it is the cheapest way to stop a Type B crew reconstructing an archive in the rain. That is when to specify it: always first, unless you already hold robust, accepted, plotted returns.