Former British Gas research engineer Ray Owen suggests recent cold weather raises issues of ground ice effects on pipes
The unusually cold December has been extremely damaging for roads and buried infrastructure, not least for the water supply network in Northern Ireland.
In south-west Scotland the winter frosts have disrupted the water supply network; villages have been replumbed.
The BBC reported workers removing cylinders of ice from pipes and that the new pipe networks were being buried at a depth even lower than the normal 750mm to minimise the risk of future deep frost.
Frost heave, where ice forms into lenses, is well documented.
Readers who think that this is a new phenomenon should perhaps look at the old Transport and Road Laboratory/HMSO publication, Soil Mechanics for Road Engineers, which details the problem quite nicely.
The size of ice lenses and the rate of formation are linked to factors such as permeability and the availability of water.
It is tempting to suggest that the ready availability of water from leaking pipe joints might just have aggravated the situation this winter.
Recent urban redevelopment and the higher expectations of customers has resulted in increasing disturbance to pipe bedding which, coupled with greater superimposed loading from traffic, has resulted in rising levels of mains failure.
In Victorian times gas and water pipes were often laid by the same company at the same time using the same cast iron pipe materials.
“Industry folklore suggested that pipe failures in the gas and water industries increased during winter months.”
Leakage from gas mains has significantly greater implications from a health and safety aspect than leakage of water and a majority of the UK gas distribution network has already been replaced in flexible polyethylene.
A research programme at the Engineering Research Station of British Gas in the 1970s highlighted the differences in the material properties of various cast iron pipes and demonstrated the significance of fissure corrosion in shortening the working life of spun iron pipes.
Full-scale testing and field trials were carried out comparing strain levels in well laid and badly laid pipes, with and without traffic loading.
Pipe strains were highly dependent on bedding conditions, poor bedding resulting in steady and damaging increases in pipe strain.
Experiments with heavy goods vehicles demonstrated that pipe strain could be quadrupled where the road surface was damaged and uneven, whilst interference from poorly supported adjacent deep excavations and tunnelling could lead to mains failure.
Industry folklore suggested that pipe failures in the gas and water industries increased during winter months.
Field trials involving the measurements of soil moisture and ground temperatures were undertaken along with some largescale loading tests of pipes buried beneath purpose-made freezing canopies.
“Potentially damaging settlements could occur at pipe bed level within a time period as short as 12 hours”
The research confirmed the time lag between surface temperature changes and changes at mains depth, also the step change in pipe strain when the ground temperature reduced to below zero degrees C at pipe depth.
Excavations in public roads during the winter of 1982 had frost penetration to almost 0.9m below ground level during an extended period of below zero air temperatures without snow cover.
In order to explain the rise in mains failures linked with the onset of freezing air temperatures the hypothesis put forward within BGC was for ‘freezing induced consolidation’.
Unpublished agricultural research and BGC data showed that with ice at the ground surface there was a reversal of the ‘normal’ soil moisture profile where moisture content increases with depth.
Soil suction measurements made with surface frost present showed a transfer of moisture / water vapour upwards from the warmer subsoil to the colder freezing front at the surface.
Potentially damaging settlements could occur at pipe bed level within a time period as short as 12 hours.