In Oxford, the assumption that ground is 'stable' collapses fast when you hit saturated sands under a few metres of alluvium. We have seen this across sites near the Cherwell and Thames floodplains, where the water table sits high and loose granular layers are common. A CPT test is often our starting point, because it gives a continuous profile of tip resistance and sleeve friction we need for the simplified procedure. For projects within the Oxford ring road, combining this with SPT drilling allows us to recover samples and verify fines content directly. BS EN 1998-5 is clear: if the groundwater is within 15 m of surface and sands are present, you assess. Oxford is not in the highest seismicity zone of the UK, but the induced stresses from a moderate event—say a magnitude 4.5 to 5 at 10 km depth—can trigger a mechanism in the loose floodplain deposits that few structural engineers think about until the borehole log shows a layer of clean sand.
Liquefaction in Oxford is a low-probability, high-consequence event. A FSL of 0.9 in a 1-metre sand lens can produce 40 mm of differential settlement beneath a pad footing.
Methodology applied in Oxford
Local geotechnical conditions in Oxford
Oxford expanded onto its river floodplains during the 19th and 20th centuries—areas that were once water meadows are now dense residential and commercial zones. The consequence is that many modern foundations sit on 1 to 3 metres of fill over natural alluvium. A standard site investigation that stops at the first stiff clay layer can miss a liquefiable sand lens entirely. We have reviewed legacy borehole records from the BGS where sands were logged simply as 'sandy clay' and later proved, by CPT, to be clean sand with a friction ratio below 1 percent. The risk is not theoretical. Differential settlements from even partial liquefaction can shear underground works, crack masonry, and render a building uninhabitable. For sites within 500 metres of a watercourse, we recommend a minimum of one CPT to refusal alongside rotary boreholes, so the stratigraphy is resolved before the foundation design is locked.
Our services
Our Oxford-focused liquefaction assessments deliver the specific outputs structural and civil engineers need for foundation design in the city's alluvial corridors. We keep reports concise and code-referenced.
Liquefaction Screening Report (BS EN 1998-5)
Full simplified procedure analysis from SPT or CPT data. Includes CSR calculation, CRR from corrected blowcount or tip resistance, MSF scaling for UK-appropriate magnitudes, and FSL profile. We deliver a table of post-liquefaction settlement estimates per depth increment.
Ground Improvement Verification
Pre- and post-treatment CPT comparisons for sites where vibrocompaction or stone columns have been specified to mitigate liquefaction risk. We quantify the increase in cone resistance and confirm the FSL has been raised above the design threshold.
Quick answers
What does a soil liquefaction analysis cost for a typical Oxford residential site?
For a site investigation that includes 2 to 3 CPT soundings to 15 metres depth, grain size analysis on key sand samples, and the full liquefaction screening report, costs typically range from £2,230 to £3,620. The spread depends on access conditions, number of tests, and whether SPT boreholes are also required for sampling.
How deep do you need to test for liquefaction in Oxford?
Per BS EN 1998-5, we assess all saturated granular layers down to 20 metres, or to the base of the Quaternary alluvium if shallower. In central Oxford, the alluvium rarely exceeds 6 to 8 metres thickness over the Oxford Clay, so 15-metre CPT soundings are usually sufficient.
Can you analyse older SPT data from a previous ground investigation?
Yes, provided the borehole logs record SPT N-values, sampler type, hammer energy ratio, and depth to groundwater at the time of drilling. We correct the raw N-values to N1(60) per BS EN 1998-5 and run the triggering analysis. If fines content data is missing from the original logs, we can take new samples from targeted depths to validate the model.