Due to the current energy scenario, technological advancements, and global economic context, it is necessary to frequently monitor the power quality supplied by utilities. The use of power quality meters in distribution systems has been increasingly adopted to prevent losses to both consumers and utilities. Furthermore, it is known that in real situations, more accurate meters are typically installed closer to the substation. In this sense, this work proposes a method for allocating power quality meters using genetic algorithms and singular value decomposition techniques to establish the position, quantity, and type of meter to be allocated, taking into account the proximity of the substation to more accurate equipment. The validation of the method was carried out in two scenarios of the 34-bus IEEE distribution system, resulting in a necessary configuration to ensure total observability of the system. Compared to other related studies that used the same system scenarios for testing, the algorithm was found to be more effective in situations where there is a limitation on the number of current measurement channels for each equipment and less effective when this restriction is not imposed.