Neutrino mass hierarchy and electron neutrino oscillation parameters with one hundred thousand reactor events

Proposed medium-baseline reactor neutrino experiments offer unprecedented opportunities to probe, at the same time, the mass-mixing parameters which govern $\nu_e$ oscillations both at short wavelength (delta m^2 and theta_{12}) and at long wavelength (Delta m^2 and theta_{13}), as well as their tiny interference effects related to the mass hierarchy (i.e., the relative sign of Delta m^2 and delta m^2). In order to take full advantage of these opportunities, precision calculations and refined statistical analyses of event spectra are required. In such a context, we revisit several input ingredients, including: nucleon recoil in inverse beta decay and its impact on energy reconstruction and resolution, hierarchy and matter effects in the oscillation probability, spread of reactor distances, irreducible backgrounds from geoneutrinos and from far reactors, and degeneracies between energy scale and spectrum shape uncertainties. We also introduce a continuous parameter alpha, which interpolates smoothly between normal hierarchy (alpha=+1) and inverted hierarchy (alpha=-1). The determination of the hierarchy is then transformed from a test of hypothesis to a parameter estimation, with a sensitivity given by the distance of the true case (either alpha=+1 or alpha=-1) from the undecidable case (alpha=0). Numerical experiments are performed for the specific set up envisaged for the JUNO project, assuming a realistic sample of O(10^5) reactor events. We find a typical sensitivity of ~2 sigma to the hierarchy in JUNO, which, however, can be challenged by energy scale and spectrum shape systematics, whose possible conspiracy effects are investigated. The prospective accuracy reachable for the other mass-mixing parameters is also discussed.