In recent years geo-environmental practitioners have experienced an increasing drive from regulators and water companies to assess risks to groundwater abstractions from turbidity that can be created by piling. There is currently no authoritative UK guidance on how to assess this risk.
Piling operations present a number of potential risks to environmental receptors if not correctly managed. These can include vibration and ground movement hazards, noise and creation of new pathways for contamination. Geo-environmental specialists are familiar with assessing risks from piling related to contamination, with reference to the Environment Agency’s 2001 guidance (EA, 2001), however this does not cover turbidity. The Environment Agency has recently commissioned CL:AIRE to update the guidance and it is understood that the revised version will refer to turbidity, but that the release date is unlikely to be before the end of 2023. Planning consents for developments in sensitive areas such as the Source Protection Zone 1 (SPZ1) of a public water supply borehole often include conditions to assess and mitigate risks to the abstraction, and can specifically require turbidity to be assessed.
Why is Turbidity Assessment Required ?
Abstractors of groundwater are required by the Drinking Water Inspectorate to regularly test groundwater for turbidity. The turbidity results are used as a marker for risks from pathogens such as Cryptosporidium and E. coli which the turbidity test does not differentiate from mineral particles. Therefore, if increased turbidity is detected the operator has to shut down the abstraction until mitigation has been implemented (Burris et al, 2020). This has significant implications for supply of water to local consumers and to the cost of water treatment. Additionally, increased turbidity can compromise the disinfection process, and where the abstracted water is treated using membrane filters then the filters can become fouled by the turbidity, resulting in replacement costs running to potentially millions of pounds. Operators of a site at which piling resulted in the shutdown of an abstraction could face significant financial and reputational liabilities.
The turbidity of water presented for disinfection must be less than 1.0 nephelometric turbidity unit (NTU), and in areas where background turbidity is elevated then water companies may apply their own more stringent criteria, which can be as low as 0.2 NTU. These are lower than the UK Drinking Water Standard of 4 NTU when supplied at consumer’s taps (DWI, 2016). For context, the image below shows water with 20 to 800 NTU. The abstracted water target is clear to the naked eye and a turbidity sensor is required to detected turbidity < ~50 NTU.
The low target values that must be achieved by the abstractor therefore present significant challenge to the risk assessor.
How to Assess Risks ?
Review by others has not identified an authoritative methodology for quantifying risks (Burris et al, 2020), however a qualitative approach can be employed. By development of a robust conceptual site model (CSM) similar to those used for contaminated land risk assessment, the potential risks can be qualitatively assessed. The principles of source, pathway and receptor creating a potential pollutant linkage are similar to those set out in the Environment Agency’s Land Contamination Risk Management guidance (EA, 2021). For the piling CSM the greatest emphasis is on the pathways and the source. The development of a scale cross-section is strongly recommended to both inform the assessment and to communicate it to regulators.
Where qualitative assessment identifies potential risks, semi-quantitative assessment can be undertaken to better understand risks and inform mitigation measures. In higher risk scenarios the CSM could be further developed with site-specific fracture details.
The primary source of turbidity during piling is mechanical action against the aquifer producing a microscopic rock ‘flour’ in suspension in groundwater, with different piling methods likely to result in different degrees of turbidity. Loss of cement fines before cement has cured is also a concern. The turbidity created will also be a function of the strata in which the piles are installed. No studies were identified that quantify the turbidity created by piling. However, qualitative assessment can quickly identify methods that are likely to create more turbidity. Continuous Flight Auger (CFA) and other rotary methods are likely to generate turbidity, particularly when operating in rock or fine grained strata, due to the mechanical action of the rotating parts abrading the rock or soil. For context, measurement of turbidity during drilling of 194mm diameter boreholes in Chalk using a tri-cone rock roller reported turbidity in thousands of NTU (maximum of 4,240 NTU) while rotary cored boreholes generated up to 452 NTU (Burris et al, 2020), although it is uncertain whether either would be representative of piling turbidity. Conversely, driven piles are expected to produce less turbidity.
Particle size of the aquifer will be important in determining extent of turbidity migration, with finer particles migrating further in an aquifer since they can be held in suspension at lower velocities and migrate through smaller pore sizes. Particle size will be largely a function of the geological strata. In a sandstone, particles formed should mainly be sand-sized since the bonds between grains will be weaker than the bonds within grains. Analysis of settled turbidity produced by tunnel boring machines in Chalk reported 80% of particles to be < 10.5 µm and 20% < 0.1 µm (Burris et al, 2020), which was attributed to the size of intact coccoliths in the Chalk (approximately 10 µm) and fragmentary material, respectively.
For turbidity to migrate beyond the source area then the groundwater velocity must be greater than the settlement velocity of the particles to keep particles in suspension. For intergranular flow the porewater velocity is unlikely to exceed settlement velocity, whereas in fractured rock the groundwater fracture velocity can exceed settlement velocity (Burris et al, 2020). In SPZ1 the groundwater velocity and gradient can exceed those under natural conditions, with both increasing nearer to the abstraction.
The lateral and vertical location of the source relative to the receptor will also be important in determining the risk. Piles installed in saturated strata to a similar depth as the abstraction intake will be at greater risk than piles that are much shallower than the intake, and risks increase with lateral proximity to the abstraction.
The scale of the project will affect the source magnitude, with both the number and depth of piles, and the duration of piling affecting the release rate of particles.
Other sources of turbidity include natural background of mineral particles in the aquifer, precipitation of solutes such as manganese and microbial contamination by bacteria and protozoa. The natural turbidity can also be affected by weather events such as intense rainfall and changes in groundwater level. Operation of the abstraction will also affect the turbidity of abstracted water. Stop/re-start cycles or changes in abstraction rate are major factors.
This is likely to be the most critical part of the turbidity assessment, since in most cases it will not be possible to change the receptor, and there will be other constraints on the choice of piling method such as ground strength, cost and contamination migration. For a pathway to be present then the source zone must be connected to the receptor by strata that have pore sizes greater than the particles produced and sufficiently high groundwater velocity. The focus on velocity is a significant variation from typical solute transport CSMs. The most likely scenario for this is karstic features or well-connected fractures in rock, with Chalk aquifers being at particular risk. It has been shown that in Chalk, groundwater velocity in fractures can exceed 2 km/day indicating potential for rapid transport of turbidity from site to the abstraction.
Where piles do not penetrate the abstracted strata and are separated from it by a fine grained stratum such as clay then there is unlikely to be a complete pathway, provided that the fine grained material is intact beneath the entire piled zone and for sufficient distance down-hydraulic gradient to protect the underlying aquifer. Whilst there is no defined minimum thickness for such a stratum to prevent migration of turbidity, confidence that the stratum will be continuous and of suitable material will increase with increasing thickness. Where an assessment is reliant on such a protective stratum then it should be supported by the proven thickness on site as well as desk-study information including off-site boreholes where available and review of other references such as BGS memoirs.
Attenuation and removal of turbidity caused by suspended sediment will be mainly by settlement of particles due to low groundwater velocity. Other mechanisms are dispersion within the aquifer and dilution at the receptor.
In most cases the receptor will be a potable groundwater abstraction which could be operated either for public supply or by a private operator. Groundwater fed surface waters may also be considered if in close proximity to the foundation works.
Once a potential pollutant linkage has been identified then qualitative assessment can be undertaken using the approach for land quality (CIRIA, 2001). Where risks are greater than Low then further assessment or mitigation will be required. Fate and transport models for dissolved phase contamination are not suitable for assessing turbidity migration, and review by others has not identified a practicable method for modelling migration of particles in fracture flow systems (Burris et al, 2020), therefore traditional quantitative risk assessment is not appropriate. Where quantitative methodologies have been proposed they are not known to have been recognised by regulators and the cost of collecting supporting data will be prohibitive for most sites. Semi-quantitative assessment based on dilution at the receptor may be appropriate however.
A cost benefit exercise will usually be required to determine whether it is more cost effective to modify the foundation solution to reduce risks or to undertake other mitigation during piling.
Foundation Design/ Re-design
Where turbidity risks cannot be addressed by risk assessment then foundation design changes may provide a lower cost, reduced timescales and more certain solution than other mitigation approaches. By altering the number, depth and diameter of piles it may be possible to terminate piles in strata overlying the aquifer and/or above the water table.
If the foundation solution cannot be changed to reduce risks then the most common mitigation measure is to undertake groundwater monitoring for turbidity during piling. Monitoring adjacent to the piled area allows any turbidity increase to be detected at the earliest opportunity. Monitoring can also exclude the site as a source if turbidity at the abstraction increases from another cause. Baseline and post-completion monitoring will also be required. Sentinel monitoring boreholes must be suitably located down-hydraulic gradient of the piled area, installed to similar depth as the pile bases and fitted with a filter pack representative of the aquifer material. An upgradient borehole is required to assess changes in groundwater flow direction and changes in background turbidity from natural causes such as heavy rainfall. The frequency and duration of each monitoring period will be site specific and should be agreed with the stakeholders at the earliest opportunity.
The site monitoring data should be complemented with turbidity data from the receptor borehole to show any seasonal trends or other events that affect turbidity. These data can also be used to inform the design of the monitoring programme, which will also need to consider lag-times and potential cumulative effects.
Where piles are installed in lower permeability strata then monitoring at the end of each day may be sufficient, whereas for piles installed in fractured rock with a short travel time to the receptor then real-time monitoring with telemetry and automatic alarms may be required. Real-time monitoring also offers the option to reduce piling rate to reduce turbidity.
For larger projects consideration can be given to scheduling piling to commence near a monitoring well so that worst-case data can be collected at the earliest opportunity.
Turbidity targets will be site-specific and will need to be agreed with stakeholders. The targets are often a defined increase relative to baseline conditions. When setting targets it is important to recognise the detection limits of the proposed monitoring instruments to ensure that the target can be detected.
Alternatives to monitoring that have been implemented including funding or indemnification for the abstractor to undertake additional treatment of abstracted water before disinfection, or abstracting turbid groundwater adjacent to the source and treating it before discharge to ground. However, these are likely to be prohibitively costly and time consuming to agree with other stakeholders and implement, even if agreement can be reached.
Assessment of turbidity risks from piling can be undertaken by qualitative assessment of source-pathway-receptor linkages based on a robust understanding of ground conditions. In many cases this will be sufficient to demonstrate that risk is acceptable without further works. Where the qualitative assessment identifies potentially unacceptable risks then the risk can be controlled by implementation of mitigation measures.
The authors thank Philip Burris for providing technical review.
Burris et al, 2020. Tunnelling, Chalk and turbidity: conceptual model of risk to groundwater public water supplies. P. Burris, C. D. Speed, A. E. Saich, S. Hughes, S. Cole and M. Banks. Quarterly Journal of Engineering Geology and Hydrogeology
CIRIA, 2001. CIRIA Report C552 ‘Contaminated Land Risk Assessment: A Guide to Good Practice’. CIRIA 2001.
DWI, 2016. The Water Supply (Water Quality) Regulations 2016, Schedule 1. The Drinking Water Inspectorate. 2016.
EA, 2001. Piling and Penetrative Ground Improvement Methods on Land Affected by contamination: Guidance on Pollution Prevention. Report NC/99/73. Environment Agency, May 2001.
EA, 2021. Land contamination risk management. Published online 8th October 2020, updated 19th April 2021.
UK WIR, 2012. Turbidity in Groundwater Understanding Cause, Effect and Mitigation Measures. Report 12/DW/1/4/5. UK Water Industry Research. 2012.
Article contributed by Tim Rolfe (Director, YES Environmental) and Craig Speed (Technical Director, Wardell Armstrong)