KEM-14 Liquefaction (settlement, damage) hazard and risk quantification module

Liquefaction is an effect which can occur in case of large intensity earthquakes (PGA, PGV) in combination which geologically recent uncompacted sandy/silty sediments at shallow depth, especially when these areas are loaded by heavy buildings or structures.
Currently, it has not been unambiguously excluded that liquefaction cannot occur as an effect of the production-induced earthquakes in Groningen. In addition, liquefaction is not part of the Hazard and Risk Assessment (HRA) model that is used to assess the seismic risk caused by the Groningen gas production.
The objective is to develop a module for the HRA model (the version developed by TNO), such it can be used to evaluate the liquefaction potential above the Groningen field. Therefore the following research questions will be addressed:

(1) Is there currently a method available to assess the occurrence of liquefaction in the Groningen region and can it be implemented as a separate module in the HRA model (the version developed by TNO)?

(2) Using the liquefaction module, most recent HRA model version and updated subsurface database, what is the liquefaction potential in the Groningen region?

(3) Is it possible that for Groningen type earthquakes (PGA’s, PGV’s) liquefaction occurs, possibly resulting in lateral spreading? If so, what effect can be expected for buildings and infrastructure in Groningen?

FULL RESEARCH QUESTION

This research project was aimed at addressing the soil liquefaction during earthquakes in Groningen for eventual inclusion in the risk assessment as an additional negative effect in the region.

RESEARCH PROJECT FINAL REPORT

The results of the research project addressing the research question stated consist of a literature review about the liquefaction potential and its relationship with building damage. The work has been focused on some soil liquefaction potential indices for soil liquefaction potential (namely LPI and LPIish). The link between LPI and LPIish and building damage has been found poor, in this respect the project has not been fully conclusive for Groningen, although certainly is a useful step forward to the understanding of the liquefaction risk in the region. The results indicate the effect of liquefaction due to man induced earthquakes in Groningen is very limited.

However, the produced results do not perfectly match the framework of the probabilistic seismic hazard and risk assessment for Groningen. It results from this study that the liquefaction potential indices considered in the project are not suitable for direct inclusions in the Groningen risk assessment and a quantitative and fully probabilistic (i.e., performance-based) approach to liquefaction, which is somewhat similar to non-linear site response analysis is required. This can be considered a useful result.

The project results have been reviewed by KEM scientific experts.

RESEARCH PROJECT EVALUATION

The project team is considered fit for the purposes of the research question and the quality of the work is good. The literature review is addressed adequately, although the choice of focusing on LPI and LPIish indices seems somewhat arbitrary and could have been better taken, considering the probabilistic seismic hazard and risk assessment for Groningen. See also here the answers to citizen questions related to this topic based on this KEM-14 and the KEM-05a project.

A follow up of this project would be to fund development of liquefaction fragility analysis for soil deposits and development of building fragility due to liquefaction for typical structures in Groningen. How much such a follow-up is a priority must be established based on the prospect abandonment of the field and the information gathered so far about liquefaction in the region (including this study), and can be matter of discussion within the KEM-panel and the KEM PSHRA-subpanel.

The project KEM-14 (and KEM-05a) have been presented in January at the KEM-DeepNL colloquium. The KEM-14 presentation in the DeepNL colloquium can be found here.