Characterising the composition of waste-derived fuels using a novel image analysis tool

- Samples representing waste-derived fuels were constructed using shredded waste components.
- 3x samples of different known composition were used.
- Image analysis was used to measure the composition of the samples.
- Determined composition was compared with the known composition.
- A strong correlation (typically r ⩾ 0.85) was observed.

An experimental study was completed using a previously developed and innovative image analysis approach, which has been applied here to shredded waste materials representative of waste-derived fuels. Waste components were collected from source-segregated recycling containers and shredded to <150 mm. These materials were then used to produce 3× samples of different composition. The samples were spread to represent materials on a conveyor belt, and multiple images of each sample were captured using 10 × 10 cm and 20 × 20 cm quadrats. The images were processed using ERDAS Imagine software to determine the area covered by each waste component. This coverage was converted into a mass using density data determined as part of this study, yielding a determined composition which was then compared with the known composition of the samples. The image analysis results indicated a strong correlation with the actual values (mean r = 0.89). The area coverage of the sample (10 × 10 cm or 20 × 20 cm) contributes to the accuracy as the dot-grid approach used with the particle size within the samples may result in components not being sufficiently monitored. This manuscript presents initial results of the application of an adapted innovative image-based method, and critically assesses how the technique could be improved and developed in the future.