Construction dust: improving industry best practice

Introduction

The High Speed Two (HS2) Air Quality Strategy^[1] aims to raise the bar on dust emission mitigation best practice standards and to share new knowledge as a legacy for future projects. HS2 aims to avoid emissions, but where this is not possible, has set key requirements to ensure emissions are reduced through control measures, ensuring that public and workforce exposure is minimised. A key concern around any construction site is the generation of dust and the short-term nuisance and long-term health effects thereof. While HS2 is ensuring all works are undertaken following Best Practical Means in reducing dust emissions it is important to continually ensure that these measures are effective through management, monitoring and mitigating against potential emissions.

One mechanism HS2 uses to ensure the effectiveness of mitigation measures on site, is by undertaking continuous monitoring of particulate matter (PM₁₀) along the site boundaries of medium- and high- dust risk construction sites. Monitors have a trigger alert level, which provides near real-time feedback to contractors when dust concentrations reach an action level, allowing for a quick response to investigate increased dust emissions and for remedial measures to be promptly implemented.

The trigger limits indicate when PM₁₀ from construction activities might be affecting local air quality. Developers should respond to breaches of the trigger threshold by stopping work immediately and ensuring best practice measures are in place before restarting.

This study was commissioned to re-evaluate the particulate matter (PM₁₀) action level of 250 μg/m³ measured as a 15-minute mean, previously recommended by Fuller and Green^[2]. This action level was based on analysis of a single construction site in 1999 and was subsequently adopted in the Best Practice Guidance (BPG)^[3] and the Supplementary Planning Guidance (SPG)^[4] for construction activities.

As used in the best practice guidance, the 250 μg/m³ trigger value (15-minute mean) is designed to protect the local population from construction emissions. Anecdotal evidence from construction sites suggest that breaches of the trigger concentration do not always have an obvious cause, and some occur outside working hours. Breaches of the 250 μg/m³ trigger value due to non-construction sources may cause undue concern to people living around the construction site and can be disruptive to construction activities.

Re-Assessment of PM₁₀ Action Level

HS2 funded King’s College London to conduct a study which examined the current levels at which dust monitor alerts are set to trigger on construction sites. The project was divided into 5 work packages:

Work Package 1: Testing PM₁₀ trigger values at construction sites;

Work Package 2: Assessing the performance of light scattering instruments;

Work Package 3: Identifying the implications for construction site monitoring strategies;

Work Package 4: Disseminating the findings of results to a wide range of stakeholders; and

Work Package 5: Providing recommendations for operational quality standards.

Work Package 1: Testing PM₁₀ trigger values at construction sites

Work Package 1 re-evaluated the particulate matter (PM₁₀) action level used in the Best Practice Guidance (BPG)^[3] and the Supplementary Planning Guidance (SPG)^[4] for construction activities. The action level or trigger concentration of 250 μg m³ measured as a 15-minute mean was based on analysis of a single construction site in 1999. King’s College London have collated a significant amount of construction specific measurement data since this early analysis was conducted allowing for a far more comprehensive study to be carried out.

The study aimed to:

Test the efficacy of the 250 μg/m³ trigger to discriminate between construction and non-construction PM₁₀ events using data from different types of construction projects using new EU reference equivalent PM measurements. The project also aimed to profile the concentration ranges that can be expected in rural, urban and roadside environments in and around London in the absence of construction through a dataset assembled from around nine construction sites.

Test the efficacy of alternative PM₁₀ metrics based on longer averaging times and also those based on incremental concentrations above the urban background. Assessment of the pros and cons of pre-scheme measurements were also included in the scope.

Pollution measurements from nine construction sites were analysed using modern EU reference instruments. As presented in Figure 1, construction sites used in the study covered a range of activities including road widening, demolition works, substation construction, and landscaping. This dataset comprises 1.8 million measurements and is the largest analysis of construction PM₁₀ to our knowledge. Construction sites are a clear source of local PM₁₀ concentrations, and many contribute to local breaches of EU limit values.

Figure 1 below shows the construction sites where pollution measurements were analysed.

This trigger limit can be used to indicate when PM₁₀ from construction activities might be affecting local air quality. It can provide important near-real time feedback to operators enabling them to take rapid and responsive measures to control emissions as part of a dust emissions control plan. The trigger is not based on any health standards and does not indicate a breach of EU Limit Value concentrations or occupational limits, merely the likely presence of a nearby construction source. This trigger will not be a perfect detector of construction emissions, but false detections should be around 0.5% of construction days.

From the nine construction sites used in the study the worst case showed the trigger value being exceeded on around one day in three. Three construction sites showed triggers being exceeded on more than one day in 12. By contrast, some sites showed no more than the expected false alarm rate. This shows that there is considerable scope for good site management practices to control construction dust. Even by controlling peak concentrations, to ensure that the trigger is not exceeded, local PM concentrations from construction might still increase by 4-5 μg/m³ as a median over the entire construction project.

The analysis of this uniquely large dataset avoids the need for pre-scheme measurements to characterise urban PM₁₀. However, care needs to be taken in rural settings where agricultural activities and local fires can give rise to exceedances of trigger values in the absence of construction activities.

The study recommended a revised trigger concentration of 190 μg/m³. This should be measured as an hourly mean.

Work Package 2: Assessing the performance of light scattering instruments

The recommendation of a trigger concentration of 190 μg/m³ measured as an hourly mean was based on measurements made using PM₁₀ instruments that have demonstrated equivalence to the European Union reference method. Practical constraints mean that smaller and simpler instruments are generally used for perimeter measurements around construction sites.

Work package 2 considered the performance of two commercially available instruments (Osiris and E-sampler). In previous short-term test these instruments required correction factors to achieve an acceptable performance envelope. For one instrument, the required correction was significantly different between tests.

For the first time, long-term data sets (up to 4 ½ year) were examined from measurement locations within, or in close proximity to construction sites. Both instruments assessed had difficulties in the measurement of volatile particulate. During pollution episodes, this volatile particulate can dominate the air in urban areas of the UK and Europe.

It was difficult to isolate construction as the source of dust from the datasets. Based on limited data there was no evidence to support a modification to the trigger concentration of 190 μg/m³ when measured with indicative instruments. In addition to difficulties in the measurement of volatile particles, even with high quality assurance control, instruments displayed long-term drift.

Indicative monitors are lower cost analysers than EU reference and reference equivalence instruments and therefore would not be expected to report data of the same quality or consistency. They are operated in rapidly changing and challenging environments that are characterised by high levels of dust and often with nearby mist cannons or water sprays used as dust mitigation. Equipment operation is often undertaken by relatively untrained staff.

The study recommended some basic steps to maximise data quality and improve site alert systems and any subsequent data analysis include:

• Good quality setting with a free movement of air around the inlet and clear lines of sight to expected sources;
• Correct configuration of instruments; paying particular attention to ensure that the sample inlet system is heated to reduce interference from water and secondary PM;
• Regular visits to change filters and adjust flows as necessary and to assess site environs to ensure that the monitor and location remain fit for purpose;
• Regular servicing, either on-site or back to base for cleaning and recalibration; and
• Regular data download and checking to ensure that equipment remains operational, to assess for consistency over time and make between instrument comparisons to identify any outlier performance issues.

Work Package 3: Identifying the implications for construction site monitoring strategies

In urban areas PM₁₀ comes from many sources including traffic, industry and wood burning. Particles also come from natural sources such as sea salt and windblown dust transported over long distances. Smaller particles tend to be formed by combustion or chemical reactions between gaseous air pollutants and can remain in the air for a week or more meaning that very distance sources can affect our air. Some particles are easily volatile making them hard to measure.

Construction sites can add to local PM₁₀ concentrations, this can come from construction machinery exhaust and also dust generated from demolition and construction activities. As shown in Figure 2, according to the London Atmospheric Emissions Inventory (LAEI) Construction accounted for 34% of PM₁₀ emissions in Greater London in 2016^[5].

The majority of construction sites assessed caused an increase in the number of days when the daily mean PM₁₀ EU limit value concentration was exceeded. There was no consistency in the type of construction activity that gave rise to high peak concentrations suggesting the need to manage PM₁₀ in all types of construction. The greatest PM₁₀ concentrations were measured very close to construction activity consistent with what is understood about atmospheric dispersion. The distance between source and receptor is important, with concentrations falling with distance. There was a considerable difference between the worst and best sites indicating the scope for good site management practices to control construction dust. Additional local PM₁₀ during the construction periods can also arise on haulage routes and local roads, most likely from the resuspension of ‘tracked out’ material.

Measurements before or after construction were compared with those during construction period. Based on this analysis an hourly mean trigger value of 190 μg/m³ (PM₁₀) was recommended for the identification of construction dust. This trigger value is applicable in urban areas in mainland UK. The new trigger provides an important lowering of the false alarm rate giving site managers and the public greater confidence that measured triggers will be due to construction activity.

Pre-scheme measurements are recommended where significant local sources could produce short term PM₁₀ peaks. Such sources include waste transfer facilities and agriculture in rural areas. Based on data assembled for this project, baseline or pre-scheme measurements would not be necessary in most urban settings.

Many practical issues restrict the use of EU reference equivalent instruments around construction sites. Smaller indicative light scattering instruments are therefore used in practice. These indicative instruments allow measurements in locations that would otherwise not be possible with EU reference equivalent methods. Their use enables the opportunity for more comprehensive measurement programmes to be conducted close to sensitive populations or close to construction sources.

Work Package 4: Disseminating the findings of results to a wide range of stakeholders

Work Package 4 aimed to disseminate the findings of work packages 1 to 3 to a wide range of stakeholders, including:

• HS2 engineering and environment teams.
• London boroughs, major construction projects and other stakeholders through the London.
• Centre for Low Emission Construction.
• Scientists at the European Aerosol Conference.
• Policy makers and scientists at the Royal Society of Chemistry AAMG annual air quality event.
• Local authorities along the proposed HS2 route.
• Environmental professionals, mainly those in the consultancy sector via the Institute of Air Quality Management.

Work Package 5: Providing recommendations for operational quality standards

The result of the study recommended a revision of the trigger levels, examined the real-world performance of the indicative instruments that are used to measure PM₁₀ around construction sites and aimed to make practical recommendations to improve the routine measurement of PM₁₀ in construction settings.

The following minimum operational requirements were made:

At the start of a sampling campaign instruments should be:

• Set up in the same manner as in the MCERTS field tests. This point applies especially to sample heaters / systems and flow
• Either
  1. Co-located at the construction site or similar location for at least one week. Reduced (or standardised) major axis regression should be used and bias between instrument pairs should be less than 10% before deployment; or
  2. Calibrated by the manufacturer or a test house with traceability to national standards.

The following actions should be carried out monthly by the field operator:

• Sample flow should be checked and adjusted with a traceable flow meter with an uncertainty of 5% or better. The field flow meter should be checked annually to ensure continued traceability of measurements.
• The operation of the heater needs to be verified by continuous measurement of sample temperature or by manual verification of heater operation. A hand-held pyrometer was found to be practical for instrument type I.
• HEPA filters should be fitted to the sample system and the measured concentration should quickly fall to within the signal noise or limit of detection for the equipment. Failure may indicate a fault which needs to be remedied.

In-service quality checks should be undertaken monthly:

• PM₁₀ concentrations should be compared during periods of low concentrations of local sources such as hours 1 to 3 each night or Sundays between pairs of instruments. Reduced (or standardised) major axis regression should be used.

1. In the event of a bias (gradient) change of ± 20%* (single month or cumulative change in bias between instruments:
   1. Equipment should be investigated for faults. If faults are found these should be remedied.
   2. If a local overnight or Sunday PM₁₀ source is found that would have interfered with the comparison, then no further action need be taken. A local source would be indicated by outliers in the regression or poor correlation.
   3. The instrument should be placed in a collocation exercise on the construction site. If a bias of greater than 10% is found, then the instrument should be subject to manufacturer or test house calibration.

Additionally:

• Instruments must be serviced in accordance with manufactures recommendations. Servicing needs include sampling systems.
• Good record keeping is essential to track the performance of instruments over time. This should include records of field checks, pre and post service calibrations and the results from in-service cross checks.

The proposed quality checks imply that spare equipment needs to be available to enable a full sampling programme to continue if equipment is withdrawn for repair or collocation checks.

Conclusions

HS2 Ltd submitted the completed study to the authors of the current national best practice for construction site dust control, the Institute of Air Quality Management (IAQM), for their consideration. After peer review, they determined the study credible to the extent that it was the catalyst to review and update their own guidance on Monitoring in the Vicinity of Demolition and Construction Sites to align with the study’s findings and adopt the new trigger action level.

This study^[6] has therefore resulted in introducing changes to environmental working procedure and updating of national best practice for the entire construction industry.

This update, specifically the Institute of Air Quality Management Guidance on Monitoring in the Vicinity of Demolition and Construction Sites^[7] has resulted in the adjustment of the trigger alert level for PM₁₀ as well as a set of quality assurance measures that should be implemented across construction sites, in line with the conclusions of this study.

Implementation on HS2

In line with the HS2 Air Quality Strategy to raise the bar on vehicle emission standards and leave a legacy for future projects, the outcomes of the study including revised trigger alert level and quality assurance measures were adopted across HS2 sites.

Chapter 7 of the HS2 Phase One Environmental Minimum Requirements Annex 1: Code of Construction Practice^[8] commits HS2 to following national best practice, and specifically cites IAQM guidance for construction site dust control and monitoring, including future revisions. Therefore, upon its publication in October 2018, all HS2 sites immediately complied with the recommendations of the new guidance. This meant adjusting dust monitor trigger levels to the new metrics and applying the monitoring QA procedures to ensure accurate data.

As a means to ensure on-site mitigation measures are effective, continuous monitoring of particulate matter (PM₁₀) is undertaken along the site boundaries of medium- and high- dust risk construction sites. Monitors have a set trigger alert level, in accordance with the study, which provides near real-time feedback to contractors when dust concentrations reach a set action level, allowing a quick response to increased dust emissions and remedial measures implemented.

Monitors are placed on or as close to HS2’s site boundaries as possible, ensuring that monitored concentrations are higher than that of relevant exposure due to dispersion, this ensuring the impacts of our construction activities are reduced, managed and mitigated. Having the action level set in accordance with the outcomes of the study has resulted in less false trigger alerts which result in work stoppages and programme delays.

The study updated national best practice for the whole construction industry, showcasing how HS2 is driving change in the Air Quality sector currently as well as leaving a legacy for all future construction projects.

Acknowledgements

• Gordon Allison (Principal Consultant at DustScanAQ)
• Susie Robinson (Principal Air Quality Consultant at Atkins)

References

1. High Speed Two Limited. Air Quality Strategy. [Accessed 3^rd August 2020].
2. Fuller, G.W., Green, D. The impact of local fugitive PM₁₀ emissions from building works and road works on the assessment of the European Union limit value. Atmospheric Environment. 2004; 38: 4993-5002.
3. Greater London Authority – London Councils. The control of dust and emissions from construction and demolition. [Accessed 1^st December 2015].
4. Greater London Authority. The control of dust and emissions during construction and demolition Supplementary planning guidance. [Accessed 3^rd August 2020].
5. Greater London Authority. The London Atmospheric Emissions Inventory 2016. [Accessed 3^rd August 2020].
6. Font, A., Fuller, G.W. Re-assessment of the 250 µg m^-3 action value. [Accessed 3^rd August 2020].
7. Institute of Air Quality Management. Guidance on Monitoring in the Vicinity of Demolition and Construction Sites. [Accessed 3^rd August 2020]
8. High Speed Two Limited. Environmental Minimum Requirements Annex 1: Code of Construction Practice. [Accessed 3^rd August 2020].