This paper presents a new methodology to calculate loss of load cost (
LOLC) indices in generating systems. It is demonstrated that all previ
ously proposed methods for evaluating LOLC indices are, in fact, appro
ximations. Sequential Monte Carlo simulation: representing all possibl
e chronological aspects, is used to calculate the probability distribu
tion of the LOLC index. It is shown how the LOLC index, expressed in d
ollars per year, and the risk of not getting it, expressed as a probab
ility or percentage, can be used in the expansion planning of generati
on systems. Both parameters, LOLC and risk, are established rt priori
as targets. Case studies with the IEEE Reliability Test system and the
Brazilian Southeastern System are presented and discussed. (C) 1997 E
lsevier Science S.A.