Norwegian Geotechnical Institute (NGI) Digital Archive
Not a member yet
1356 research outputs found
Sort by
Evaluation structures for machine learning models in geotechnical engineering
Full text linkpublishedVersio
Machine learning coupled with causal inference to identify COVID-19 related chemicals that pose a high concern to drinking water
Full text linkpublishedVersio
A laser ultrasound system to non-invasively measure compression waves in granular ice mixes
Full text linkpublishedVersio
PFAS, PCBs, PCDD/Fs, PAHs and extractable organic fluorine in bio-based fertilizers, amended soils and plants: Exposure assessment and temporal trends
Full text linkBio-based fertilizers (BBFs) produced from organic waste contribute to closed-loop nutrient cycles and circular agriculture. However, persistent organic contaminants, such as per- and poly-fluoroalkyl substances (PFAS), polychlorobiphenyls (PCBs), polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs), as well as polyaromatic hydrocarbons (PAHs) can be present in organic waste or be formed during valorization processes. Consequently, these hazardous substances may be introduced into agricultural soils and the food chain via BBFs. This study assessed the exposure of 84 target substances and extractable organic fluorine (EOF) in 19 BBFs produced from different types of waste, including agricultural and food industrial waste, sewage sludge, and biowaste, and through various types of valorization methods, including hygienization at low temperatures (140 μg kg−1), so monitoring of more PFAS is recommended. The calculated expected concentrations in soils after one BBF application (e.g. PFOS: <0.05 μg kg−1) fell below background contamination levels (PFOS: 2.7 μg kg−1) elsewhere in the literature. This was confirmed by the analysis of BBF-amended soils from field experiments (Finland and Austria). Studies on target legacy contaminants in sewage sludge were reviewed, indicating a general decreasing trend in concentration with an apparent half-life ranging from 4 (PFOS) to 9 (PCDD/Fs) years. Modelled cumulative concentrations of the target contaminants in agricultural soils indicated low long-term risks. Concentrations estimated and analyzed in cereal grains were low, indicating that exposure by cereal consumption is well below tolerable daily intakes.publishedVersio
Emission Factors for Biochar Production from Various Biomass Types in Flame Curtain Kilns
Full text linkpublishedVersio
Deliverable 4.6 Design of multi-site and multi-scale monitoring schemes
Full text linkThis report describes a monitoring scheme applicable for CO2 storage projects at different maturity levels and scales. It is the final delivery from ACT SHARP (“Accelerating CCS Technologies - Stress history and reservoir pressure for improved quantification of CO2 storage containment risks”) Work package 4 (WP4) – “Monitoring”. It builds on previous deliveries from WP4 and other SHARP work packages on multi-scale geomechanical rock failure risks (both onshore and offshore), machine-learning approaches for seismicity detection, and optimal use of fibre optics.
In this report we introduce the concept of Geomechanical Readiness Level (GRL), a scale intended to help storage operators evaluate the readiness of their potential injection site with respect to available data characterising the stress conditions at the site. We discuss how selected potential storage sites in the North Sea and onshore India place in the GRL scale, and how work performed both within SHARP and by individual operators has matured each site to their present GRL level.
We also present an overview of monitoring tools for detecting geomechanical pore pressure and stress changes and discuss their different applications. The focus in SHARP has especially been on potential fibre optics applications, which have the potential to detect a variety of subsurface changes (pressure, temperature, seismic responses, strain), but which also comes with limitations (directivity, deployment limitations, noise, handling large data volumes). An important delivery from WP4 has been developing efficient data analysis tools, capable of combining observations from different sources and handling large datasets with machine learning.
Finally, we describe how an optimal monitoring programme needs to be tailored to the site(s) in question, selecting optimised monitoring tools depending on the relevant risks and geomechanical setting.European CommissionpublishedVersio
Unsupervised Machine Learning for Data-Driven Rock Mass Classification: Addressing Limitations in Existing Systems Using Drilling Data
Full text linkpublishedVersio
Geotechnical laboratory testing of lunar simulants and the importance of standardization
Full text linkpublishedVersio
