Parallel computing of sequential MonteCarlo techniques for reliable operation of Smart Grids

Probabilistic methods based on Monte Carlo techniques are successfully employed, both in vertically integrated utilities and in deregulated markets, to assess the long-term adequacy of power systems. Unfortunately, the computational complexity of many time-consuming algorithms has constituted for years a crucial barrier to the use of probabilistic methods also for short-term reliability assessment and decision-support. Nowadays, the availability of cheap and fast computers discloses new opportunities also for on-line reliability calculations. In particular, parallel computing seems to have good chances to be successfully applied to the on-line operation of Smart Grids. In the present paper, a technique of parallel computing, applied to a sequential Monte Carlo algorithm that simulates a power system in a multi-core machine, is proposed and described. A case study based on the IEEE Reliability Test System RTS-96 is shown and discussed, with particular emphasis to the speed-up and accuracy obtained as a function of number of employed cores.