The distributed virtual forces deployment algorithm provides a very good area coverage and guarantees network connectivity for a sufficient number of sensor nodes. It relies on local information between neighboring sensor nodes. However, its main drawback is node oscillations and hence a high amount of sensor node energy wasted. The grid based strategy divides the monitored area into virtual cells. Each cell center determines the position of a sensor node. In this paper we propose GDVFA that combines the advantages of both strategies: on the one hand coverage and connectivity for virtual forces strategy and on the other hand avoidance of node oscillations for grid strategy. Simulation results reported in this paper show that GDVFA considerably reduces the energy consumed by sensor nodes. This comes from: 1) the detection of redundant nodes that are put in sleep mode, and 2) the avoidance of node oscillations by stopping nodes.