[2025] UKUT 053 (LC)

The positions of level monitors and the maximum reading at each are summarised in the relative movement sketch below. The first number is the point number and the second number is the greatest level difference in mm recorded at that point over the period. Two points are missing: point 4, which is close to point 3 and showed very similar readings, and point 9, which is close to point 8 and showed very similar readings. Visually it is apparent that the greatest movement was to the rear of the garage at points 10 (-27mm), 11 (-32mm) and 12 (-24 mm). The only significant movement on the extension is shown at point 6 (-21mm). It will be noted that point 6 is the corner of the extension where the TPBH1 investigations were made.

The level monitoring results over the period for points 6 – 10 and 11 - 12 are shown in the charts below, with labels added as an aid. It should be noted that the y axis scale is different on each chart to accommodate the relevant range of movement. These charts were analysed by the experts to look for evidence of seasonal and cyclical movement, which would be indicative of tree root desiccation of shrinkable clay, so particular note was taken of any recovery over the winters 2015-16, 2016-17 and 2017-18.

The engineering experts agreed that the monitoring at points 10, 11 and 12 around the garage displayed a seasonal pattern of movement indicative of tree root desiccation of a shrinkable clay subsoil below the foundations, but Mr Freeman considered that this could have been caused by any of the trees at the rear of the property, including the cypress tree and privet hedge growing close to the garage. There was obvious progressive downward movement recorded at points 6, 10, 11 and 12 and Mr Freeman considered that the soil at the rear had been affected by sub-surface erosion. The moisture contents recorded in BH1 at depths of 0.9m and 1.5m were unnaturally high and indicative of a soil that had undergone swelling and softening, which can only occur where a void is created by excavation or erosion. The presence of the soft layer was confirmed by the suction tests and the exceptionally low shear vane reading of 39kPa at 0.9m. In his view the only plausible explanation for the progressive downward movement was sub-surface erosion caused by water escaping from the leaking drainage and running down the slope under the foundations to the extension and the garage. Softened soil was a symptom and it was compression of the voids that would cause downward movement to the supported foundations. Drainage repair in autumn 2015 may have encouraged erosion by lowering the water table, allowing a higher flow of water through the soil in the summer. No investigation was made in June 2015 of the downpipe and drains at the rear of the garage to rule out leakage there too.

In Mr Allen’s opinion, whilst the high moisture content was likely to be the result of nearby drain failure, the subsoil still had adequate strength to resist the loads applied by the foundations. 39kPa equated to a safe bearing pressure of 65 kN/sqm, which exceeded the estimated load applied by the foundation of 40kN/sqm. London clay soil would swell if it had been previously desiccated and there was no evidence of void formation below foundation level in the BH1. The sub-surface erosion mechanism described by Mr Freeman would have been present since the property was constructed and not commenced in 2014. In his opinion the reduced amount of upward movement, compared with subsidence, resulted from a persistent moisture deficit in the clay subsoil where winter rehydration was less than the desiccation of the previous summer. Once the drains were repaired in 2015 the roots were likely to have been desiccating to a greater depth than recorded in the 2015 site investigation.

The arboriculture experts also considered the level monitoring results. In Dr Dobson’s opinion the high moisture content of the soil around TPBH1, in June 2015 when the drains were found to be leaking, probably caused oak roots to proliferate and then continue to extract water after the drains were repaired in September 2015, causing progressive desiccation and downward movement superimposed on the general pattern of subsidence and recovery. He considered that the limited upward recovery in the winter of 2016-17 may have been because of drier than average conditions in October and December 2016, and again in early 2017, limiting soil rehydration.

It was Mr Pryce’s opinion that development of a persistent moisture deficit would require a high density of active live roots, which would tend to occur close to the trees rather than at the periphery of the root system. T1 and T2 were mature trees so their rooting zones would be unlikely to have altered appreciably in recent years to have caused the downward movement. Mr Pryce agreed that roots would proliferate where conditions were favourable, but considered those found in TPBH1 were outliers that would have become redundant once the leaking drains were repaired. The trees (T1 and T2) had good rooting conditions in every other direction so would have stopped supplying these roots with their energy supply of starch, in favour of roots closer to them with easier sources of water.

Dr Dobson referred to reported research in which monitoring of ground level on shrinkable clay near to an oak tree had been carried out since 2006 by Mr Freeman’s company on behalf of the Clay Research Group. Subsidence movements were found to be greatest at a distance of 1.2 times the height of the tree and Mr Freeman had commented that the zone of influence of a larger tree could extend beyond 1.2 tree heights away. In this case where T1 was 28m tall, the zone of influence could therefore extend beyond 33.6m, which was further than the distance to the garage and the extension. Dr Dobson reiterated that oak roots were present under the foundations of the extension, and although only a few were found in the borehole, it was only 70mm in diameter and they must have been plentiful.

Mr Pryce considered that measures of tree height to influencing distance were convenient in providing a guide for risk management exercises but were not reliable as an indicator of root spread in individual trees. From published research, the maximum distance at which oak tree roots had caused damage was 30m, and then for only a small fraction of the tree population. He acknowledged, fairly, that the presence of roots meant there was a possibility that the oak tree roots had spread sufficiently to be the cause of damage but, in his long experience as a tree inspector, this seemed unlikely.