This paper investigates the operation of New Zealand’s electricity system replacing large thermal generation with different proportions of wind and geothermal generation. These portfolios are assessed with an implicit stochastic optimisation modelling tool that uses dynamic programming with backward recursion as the deterministic optimisation, and the forward simulation conducted with a weighted linear regression. The shortfall in generation capacity is identified for a range of generation and inflow scenarios, and the unserved energy is calculated. As the development of the modelling tool is still in progress, the modelling tool and specific results presented in the paper are not final and will differ from those presented. However, the conclusions drawn from the results will likely be the same.