Abstract. We consider the detector size, location, depth, background, and radio-purity required of a mid-Pacific deep-ocean instrument to accomplish the twin goals of making a definitive measurement of the electron anti-neutrino flux due to uranium and thorium decays from Earth's mantle and core, and of testing the hypothesis for a natural nuclear reactor at the core of Earth. We take the experience with the KamLAND detector in Japan as our baseline for sensitivity and background estimates. We conclude that an instrument adequate to accomplish these tasks should have an exposure of at least 10 kilotonneyears (kT-y), should be placed at least at 4 km depth, may be located close to the Hawaiian Islands (no significant background from them), and should aim for KamLAND radio-purity levels, except for radon where it should be improved by a factor of at least 100. With an exposure of 10 kT-y we should achieve a 25% measurement of the flux of U/Th neutrinos from the mantle plus core. Exposure at multiple ocean locations for testing lateral heterogeneity is possible.