Rock stability considerations for siting and constructing a KBS-3 repository. Based on experiences from Aespoe HRL, AECL's URL, tunnelling and mining

Over the past 25 years the international nuclear community has carried out extensive research into the deep geological disposal of nuclear waste in hard rocks. In two cases this research has resulted in the construction of dedicated underground research facilities: SKB's Aespoe Hard Rock Laboratory, Sweden and AECL's Underground Research Laboratory, Canada. Both laboratories are located in hard rocks considered representative of the Fennoscandian and Canadian Shields, respectively. This report is intended to synthesize the important rock mechanics findings from these research programs. In particular the application of these finding to assessing the stability of underground openings. As such the report draws heavily on the published results from the SKB's ZEDEX Experiment in Sweden and AECL's Mine- by Experiment in Canada. The objectives of this report are to: 1. Describe, using the current state of knowledge, the role rock engineering can play in siting and constructing a KBS-3 repository. 2. Define the key rock mechanics parameters that should be determined in order to facilitate repository siting and construction. 3. Discuss possible construction issues, linked to rock stability, that may arise during the excavation of the underground openings of a KBS-3 repository. 4. Form a reference document for the rock stability analysis that has to be carried out as a part of the design works parallel to the site investigations. While there is no unique or single rock mechanics property or condition that would render the performance of a nuclear waste repository unacceptable, certain conditions can be treated as negative factors. Outlined below are major rock mechanics issues that should be addressed during the siting, construction and closure of a nuclear waste repository in Sweden in hard crystalline rock. During the site investigations phase, rock mechanics information will be predominately gathered from examination and testing of the rock core and mapping of the borehole walls. Two major tasks must be accomplished during this period: 1. an assessment of the quality of the rock mass and 2. an assessment of the state of stress within the volume of rock containing the repository. Empirical methods such as the Q system can be used to establish the domains of rock mass quality and to assess tunnel support requirements during the preliminary design phase. The laboratory testing should be carried out to determine the crack initiation stress, the long-term strength, peak strength and post-peak response. The determination of these parameters should be determined from stress-strain data, as well acoustic emission testing techniques, using testing methods based on accepted national standards, such as the ISRM suggested methods or ASTM. The in-situ stress state must be measured with confidence. The number of measurements and the method(s) used will be a function of the geology of the site. Practical experience indicates that stress-induced failure (spalling) will occur on the boundary of an underground opening in hard rocks when the maximum tangential stresses on the boundary of the opening exceed approximately 0.3 to 0.4 of the laboratory uniaxial compressive strength. Hence to assess the potential for spalling, numerical analysis will be required for the various shaped openings planned for the repository. These numerical analysis can be used to optimize the shape of the tunnels, the orientation of the tunnels relative to the far-field stress state, intersection support, and deposition tunnel/borehole spacing. The support for the tunnels in a repository is expected to range from light support pressure equivalent to standard spot-bolting to local bolts with mesh and fibre-reinforced shotcrete. At major intersections medium to heavy support pressure may be required. The layout of a repository will be similar to a mine using a room-and-pillar mining method but the extraction ratio will be of the order of <30%. A drill-and-blast excavation method will provide the maximum flexibility for such an excavation technique. The two common modes of failure (structurally controlled and stress-induced spalling) can be analyzed using the approaches outlined in this report. However, it is not clear what approach should be used when the mode of failure is transitional, i.e., structure/stress. Apart from the ZEDEX experiment, very little rock mechanics research has been carried out with the combined in-situ stress magnitudes and well defined structure such as occurs at the 400 to 450 Level in the Aespoe HRL. Because of the likelihood of encountering these transitional conditions at the depth of the proposed repository in Sweden it is recommended that further rock mechanics research be carried out to assess the stress at failure for these transitional conditions. In particular, the strength of the pillars between the emplacement boreholes should be established such that the pillar dimensions can be assessed by means other then empirical formulas developed from mining conditions.