Patterns of ground movement have been characterised for a variety of perils including (a) seasonal movement associated with root induced clay shrinkage, (b) heave, (c) progressive movement of the sort caused by leaking drains, poor ground and (d) the absence of movement or change – i.e., stability.

Diurnal temperature change and interference (loss of power for example, or loss of signal) are removed from the analysis. If the data shows change, then we can be sure it is due to movement or moisture variations.

Stability is resolved in a similar way. The above values are known and any recorded change less than the known values are discarded, indicating stability.

The application has been developed primarily to assist in the detection and diagnosis of root induced clay shrinkage. Although it offers probability values, the final decision on causation rests with the engineer.

The engineer has to use the data in conjunction with an assessment of the damage, timing of notification and understanding of the geology before arriving at his decision. The application is meant as a support tool and no liability attaches to its use.


This will vary but have also assumed that the use of telematics and sophisticated sensors implies the claim warrants this level of attention. The duration of monitoring may therefore extend over a considerable period of time. Conversely, because the application detects small traces quickly, where appropriate and supported by the engineers judgement, it may be possible to significantly reduce the monitoring term and below we offer some guidance.

Clay shrinkage claims can be identified in a very short period of time. The ground is recovering (getting wetter) from September through until December which could be a water problem, but when combined with recovery of the electrolevel, then it is a positive indicator. As clay is the only soil that exhibits this seasonal behaviour, we can identify root induced clay shrinkage positively in a few months with the caveat that heave exhibits similar characteristics and here the diagnosis will be swayed by the presence (or absence) of vegetation and crack patterns.

Contraflexure of both building movement and moisture change in September is a positive indicator of root induced clay shrinkage.

In these cases we would recommend two months as being the minimum period to remove any ‘background noise’ related to localised rainfall and temperature change etc.

Any pattern that doesn’t correspond to the periodic signature of seasonal clay shrinkage is, by definition, ‘something else’ and that could include heave, water or poor ground.

If the moisture content is increasing between May and September, or the active station simply wetter than the datum then it is a water problem. If it is increasing slowly, pausing in the summer before continuing then the most likely cause is heave.

Reference to the sector data in respect of the “Clay Shrinkage Risk” may be a useful distinguishing feature. Heave takes place on a clay soil, whereas water is usually troublesome in non-cohesive ground.

Reference to the datum is essential. There has to be a difference between this and the active station to arrive at a decision. Where there is no difference, or if there is, but there is no change over time, then we have stability. Ideally this might extend over 12 months depending on circumstances, but certainly over a term used to diagnose root induced clay shrinkage if just to rule out a seasonal pattern.


Heave takes longer to diagnose and we recommend a full 12 months to detect the flattening in the summer months, prior to onward upward movement. As the rotation associated with heave causes complex failure patterns, electrolevels and moisture probes are needed. The indicator is ongoing recovery and an increase in moisture that halts in the summer.

It is vital that the engineer carries out a full assessment of the damage and site conditions before making a diagnosis.

Clay Shrinkage

This is characterised by...

1. Ground drying between May and September.

2. The difference between the datum (movement and moisture) and the station increasing towards the summer and decreasing towards the winter.

3. Contraflexure in September for both movement and moisture.

4. The ‘Clay Shrinkage Risk’ is usually in the ‘medium’ or ‘high’ risk category.

Escape of Water/Poor Ground

This is characterised by...

1. Gradual and ongoing downward movement, possibly at a steeper rate than the heave profile.

2. The ‘Clay Shrinkage Risk’ is usually in the ‘low’ risk category.

Escape of Water

This is characterised by...

1. An increase in moisture content

2. Wetter at the point of movement than at the datum.

3. Downward movement between September and May.

4. The ‘Clay Shrinkage Risk’ is usually in the ‘low’ risk category.


This is characterised by...

1. A gradual increase in moisture content and movement, slowing or even reversing slightly between August and September, before continuing.

2. The slope of the line is usually slower than for poor ground and escape of water.

3. The ‘Clay Shrinkage Risk’ is usually in the ‘high’ risk category.

Poor Ground

This is characterised by...

Often follows a similar pattern to escape of water, and can be exacerbated by an increase in moisture – slippages often occur early in the year or following heavy rainfall for example.


The application uses product moment correlation technique to measure the relationship between data sets that are scaled to be independent of the unit of measurement. For example in our analysis we compare ground movement (mm) with tilt (degrees) and moisture change (expressed volumetrically as a percentage). The algorithm uses pattern matching, rather than absolute values.

The application returns the covariance of two data sets divided by the product of their standard deviations. The actual readings are matched against a library of ‘charactersitic signatures’ representing the typical profiles for ground movement for the various categories of damage.

‘1’ represents an excellent correlation, ‘0’ (or a negative value) tells us there is no correlation at all, and in the intermediate range, anything above 0.6 is interesting and above 0.8 generally suggests the correlation is significant.


We have provided guidance of both claim numbers and geology which is based on investigations and should be regarded as no more than a rough guide. Significant variations can exist within a postcode.


OSCAR is a statistical, web based application that models the relationship between climate, trees and geology. It can provide guidance on the risk posed by trees of varying species, height and distance from a building in relationship to the geology, and at different times of the year. It allows ‘what if’ modelling relating to climate and a predictive capability in respect of future tree growth and the likely risk expressed as an estimate of ground movement.