PhD opportunities

On-Demand History Matching in Reactive Transport. Application to U-ISR.

Thesis proposal

Area of expertiseGeoscience and Geoengineering
Doctoral SchoolGRNE - Gosciences, Ressources Naturelles et Environnement
SupervisorM. Herv CHAURIS
Research unitGeosciences
ContactCHAURIS Herv
Starting dateOctober 1st 2020
KeywordsHistory Matching, physics-based and data-driven optimization, reactive transport, ISR
AbstractIn situ recovery (ISR) has recently become the main mining technique for uranium production: it is used for roll-front deposits occurring in medium deep aquifers. Compared to conventional techniques (open pit and underground mining), ISR is faster to implement, less expensive and offers reduced environmental footprint. However, as for oil&gas field, ISR has not a direct access to the deposit within the reservoir and suffers from strong uncertainties about the initial estimate of reserves and the assessment of their evolution over time.
For the past ten years, the Center of Geosciences of MINES ParisTech and ORANO Mining have been developing a deterministic approach to simulate ISR operations using reactive transport code HYTEC. The model is based on a 3D geological model (porosity/permeability maps and distribution of reactive mineral phases) coupled with a geochemical model describing the interactions between the leaching solution and the mineral phases. In addition, the geometry of the well-field (coordinates and screen position) is fully described as well as operating scenarios.
The feasibility and robustness of reactive transport modelling for ISR were demonstrated, especially at KATCO, a mining site in Kazakhstan. It has been shown that HYTEC accurately reproduces uranium recovery at the technological block scale (about 15 producer wells and 60 injectors) with manual calibration of the direct model performed on a few geochemical parameters. However, large discrepancies may remain, particularly when analysis is made at the scale of the individual producer well. At this scale, improving the history matching results requires local adjustments of geological/geochemical models through optimization methods.

This PhD thesis aims mainly to reinforce the prediction capacities of simulations by integrating the automatic deterministic solution of the inverse problem to improve the history matching of the producer well parameters in operation area. Indeed, the forecast of production scenarios can be made at different scales, but the set of producer wells is the representative scale for the operator and the precision at this scale is then crucial for the deployment on site. The thesis has four principal stages.
1. Methodology. Development of the inverse problem using determinist methods. Integration either directly in the platform HYTEC or externally in a meta-layer using existing libraries. Development of different regularization approaches. Evaluation of using response surfaces to accelerate geochemical calculations. This approach might speed up the convergence or/and replace the geochemical calculations of ISR.
2. Characterization of major ISR parameters, i.e. the most influential in the ISR modelling. These parameters cover different aspects. For example, transport properties (porosity, permeability), distribution of geochemical facies (oxidized, mineralized and reduced zones of roll front), and geochemical model (distribution of acid consumers, distribution of Uranium grade).
3. A set of applications with increasing complexity: history matching of real production at the scale of the producer well; production forecast for mining plan; estimation of production potential; and reconciliation between production and initial estimate of reserves.
4. Comparison with stochastic methods. Collaboration with external experts is possible. For example, optimization methods in oil&gas domain.
ProfileSelection criteria
- Strong motivation for the project
- Experience in programming and development of numeric methods (applied mathematics). Desirable: optimization methods, stochastic methods.
- Excellent ability and avidity to learn and develop new methods in the different scientific fields.
- Excellent ability and avidity to model physical and chemical processes
- Excellent written and verbal communication skills in English
FundingConvention CIFRE