UNcertainty reduction in monitoring methods for Improved CO2 QUantity Estimation (uniCQue)
2014 – 2017
The primary objective of the "uniCQue" project is to develop methods for quantification and reduction of uncertainties in CO2 monitoring. The focus is on uncertainty estimation in results obtained using imaging methods such as Full Waveform Inversion (FWI) and Controlled Source Electro-Magnetics (CSEM) inversion. This is achieved by setting up a work bench of prototype codes based on existing software at SINTEF Petroleum Research. Different uncertainty quantification techniques are developed and tested in the work bench. The monitoring methods are then optimized in order to minimize uncertainties and improve CO2 quantity estimation. Tests are performed on synthetic data and on data from Sleipner.
Accurate and reliable monitoring methods are crucial for early detection of leakage and reduction of risks associated with CO2 injection and storage. World-wide efforts have been made to study and improve the accuracy of these methods, but so far very little has been done to quantify the uncertainty in the information provided by the monitoring images. Sound knowledge of the uncertainty in an image is a vital component in quantification of risks during injection. In addition, this increases the confidence in the assessment of a storage site prior to injection.
One of the main challenges related to estimating uncertainties is the computational cost. A major effort has to be made to develop efficient, yet accurate, algorithms for such calculations. Presently, very little work has been done on estimating uncertainties for methods like FWI and CSEM and there is likely a strong need for innovative solutions. Quantification of CO2, on the other hand, is an active area of research, but will also be challenging, mainly when dealing with real data.
The first work in this project included tests of a method to quantify CO2 using CSEM constrained by structural details assumed to be obtained using FWI. This work was presented at GHGT-12 and published in Energy Procedia. A Python work bench for monitoring methods was established and has since been gradually improved. A postdoc within the project recently added Mayavi visualization support to the codes. Another postdoc has worked part time in the project on how to implement acoustic wave propagation efficiently in Python. In 2015, the testing of a priori covariance analysis for uncertainty quantification resulted in a presentation at the SIAM Conference on Mathematical and Computational Issues in the Geosciences. Further work on this topic (partly done by the postdoc) will be presented at GHGT-13. In parallel with the uncertainty studies, SINTEF has been working together with BGS on improving monitoring with FWI by including information from spectral decomposition and AVO studies. The results of this work will be presented at GHGT-13.