In the context of the introduction of renewable energies in France, Nuclear Power Plant Operations (NPPO) are a key component for the compensation of the intermittent production of solar and wind power. In this work, we focus on the optimization of the operation cost and stability of power of a real-life power transient, while maintaining safety standards. From an optimization point of view, the NPPO problem is a typical example of a discrete constrained bi-objective problem based on time expensive computation simulation. We propose a massive asynchronous parallel master/workers MOEA/D assisted by a surrogate models. The algorithm design components are discussed and argued in this work. We show that our proposed surrogate assistance is able to improve algorithm performance and reliability, allowing us to extend our approach to a large range of strategic future real-life operations.