Comments of the German Association for Repository Research (DAEF)

Till Popp, Wolfgang MinkleyInstitute for Geomechanics (IfG)7th US/German Workshop on

Salt Repository Research, Design and OperationWashington, DC

September 7-9, 2016ABSTRACT

On 30.11.2015 the article, “Deformation-assisted fluid percolation in rock salt” by Ghanbarzadeh, S. et al. was published in the journal Science. This item was taken up in Germany by the media and was presented as new technical knowledge which may contradict the suitability of salt as a host rock for radioactive waste. The DAEF prepared a brief statement with the present opinion, to discuss the various theses and the reliability of conclusions made in the above-mentioned paper. The main outcome of the DAEF-paper is presented as part of discussion in the breakout session.

Firstly, as a general statement, we believe, based on our long-lasting experience of research on salt, that it’s a fact that salt is usually tight. Thus, salt is the most suitable geological barrier for radioactive waste repositories, as just might be demonstrated by the occurrence of hydrocarbon accumulations in salt. The tightness of salt results mainly from its low porosity (and permeability) generated during burial diagenesis. Due to (e,g. dislocation or humidity-assisted) salt creep, favored by increasing temperature and pressure with depth mechanical stresses are nearly isostatic and, therefore, acting external fluid pressures are usually lower than the minimal stress, which is a prerequisite for salt barrier integrity.

From the engineering point of view the processes and conditions, where salt can lose its integrity, described as percolation threshold, are well known. Two mechanisms are assumed as relevant for performance assessment (Minkley & Popp, 2010):

(1) deviatoric stresses inducing growth and connection of inter-crystalline pores as well as transcrystalline cracks – assessed by the dilatancy criterion; and

(2) fluid pressures causing hydraulic opening of cracks and grain boundaries and their interconnection – in practice assessed by the minimum stress criterion.

Here a third loss-of-tightness mechanism for salt is suggested based on the static pore-scale theory. The thesis is that salt will become permeable at greater depth (e.g. lower than 3000 m depth). The scientific basis results from the work of Lewis & Holness (1996), who observed the formation of topological connected brine-filled pores and triple-junction pores in halite/water aggregates at respective PT-conditions. As a pity, from our knowledge, no natural salt from that depths was investigated confirming the textural observations made on synthetic salt.

Thus, the first question arises, “Are the realized experimental test conditions, relevant for natural salt?”

As a fundamental critique, the authors used only small synthetic samples (around 150 mg table salt with uniform single halite-crystals, whereby 7 - 15 mg water is added). This corresponds to a saturated water-filled porosity in the order of 7 to 15% which is quite high and unrealistic for natural salt conditions.

The second question is “How can we explain the occurrence of hydro-carbons in salt?”

The answer could be very simple; hydrocarbons are a typical feature of salt formations, due to formation of organic-rich evaporitic mudstones during burial diagenesis. Therefore, it has to be demonstrated that the observed hydrocarbons have a different stratigraphic origin than the surrounding salt.

The third comment is that “The main statements about high permeability of salt rocks are derived only from pore-space simulations, and not supported by real permeability measurements”.

IfG performed preliminary measurements on natural salt cores in the relevant PT-field (p = 95 MPa, T = 120°C). We were not able to measure any gas flow (injection pressure up to 10 MPa) but the measurements need to be verified.

At the state of knowledge we believe that the impact of the dihedral angle on salt integrity at repository conditions may be low, but we suggest additional investigations to improve the understanding of these processes and to solve the uncertainties, i.e.

(1) to repeat the hot-pressing tests with lower water contents, until the order of <1% and to perform the respective texture investigations. There is probably a fluid content threshold where – due to missing water - no connecting pores may be generated, independently from the PT-conditions (typically existing in a repository for radioactive waste). This would allow solving the debate about natural salt permeability.

(2) to perform permeability tests on natural salt samples at the respective PT-conditions.