Douglas_Reay comments on Optimum number of single points of failure