Case Study 2

Site Management

The need to develop the site quickly meant that many operations such as piling and roadway construction were carried out simultaneously, with many of the activities generating excavated soil. Due to the level of activity and quantities of soil being excavated, rapid assessment and removal of stockpiles was needed as there was little space to store stockpiles while laboratory data was obtained.

SI data indicated that some of the soil excavated could be hazardous under H14 the ecotoxicity classification in the new waste classification regulations. On site analysis was therefore used to assess and segregate soil into hazardous and non hazardous as the difference in disposal costs would be around £1200 per lorry load.

Before the on site analysis was started, a waste classification model was used to estimate the concentrations of the target metals that would classify the soil as hazardous. This model showed that for the selected metals, the X Ray fluorescence technique had sufficiently good detection limits to meet the data quality objectives.

Classification and segregation of soil excavated and placed into stockpiles

The following example shows how on site analysis using an XRF analyser can be used to minimise the costs of disposing of the soil from a spoil pile contaminated with various levels of heavy metals by segregating the hazardous and non hazardous material for subsequent disposal off site.

Site History

This large site was former offices, car parking, warehousing and light industrial premises that is being redeveloped into a large retail and office space complex. The site is in the middle of a built up area on ground that has had buildings and other activity on for over 200 years, making the ground a complex mixture of soil and sand containing up to 50% brick, concrete, glass, ceramic and wood.

Initial SI data showed the main pollutants were heavy metals with some hydrocarbons in localised areas. The main metals of concern were Pb, Zn and Cu.

Summary Results

The XRF results for each sample shown in the table below are the average values from all the XRF readings taken for the soil making up that stockpile face. The MCERTS results are the laboratory values from the composite sample from the same face as the XRF data. The %>5mm is the percentage of stone and other solid material in the bulk composite sample that would not pass through a 5mm sieve in the laboratory. This value provides an estimate of the soil homogeneity of that area, with a high percentage showing very inhomogeneous material.

On Site Analysis Approach

Heavy metals pollution is often found in localised areas as the metals are usually in an insoluble form and remain where they were originally deposited. In made ground, the soil homogeneity is very poor, so samples taken from just 1 metre apart can produce very different results. This site had soil that contained lots of brick, stone, ceramics and glass (up to 50% would not pass through a 5mm sieve) which made taking representative samples for both laboratory and on site analysis difficult.

To reduce the uncertainty in the results and to attempt to get a representative metals concentration for the stockpiles, a sampling frequency of an XRF reading every 2 - 3 metres was used. The soil was pressed down in situ and a reading taken. If the result showed the area non hazardous, the spot was marked as OK and the next 2 metre point analysed. If the result was hazardous, a second XRF reading would be taken a few centimetres from the first reading. If the result still showed hazardous levels, the location was marked as hazardous.

This sampling strategy was carried out across the entire face of the stockpile. Areas that showed hazardous spots next to each other had a 1 metre depth removed up to the next OK spots and put onto a hazardous stockpile. Areas that had virtually all OK marks with no adjacent hazardous marks also had a 1 metre depth of soil removed, which was placed directly into a waiting lorry for offsite disposal. Composite samples from approximately every 4 - 6 lorry loads of the identified non hazardous soil were also taken for laboratory confirmation.

Once each face was cleared to a depth of 1 metre a new sampling run would begin on the exposed face and the analysis and soil removal repeated until all the stockpile was sorted.

The sampling average was approximately 1 measurement per 4 cube of soil, with around 20 measurements per face. This took approximately 25 - 35 minutes. Including the removal of hazardous areas and profiling of the stockpiles the rate of sorting and removal was approximately between 300 and 400 cube per day.

Most stockpiles had around 75% - 85% of their mass classified as non hazardous, providing huge savings for the client who without on site analysis would have had to send all the soil to a hazardous landfill.