Although most research in lattice wireless sensor networks are focus on determining the optimal deployment pattern to provide full sensing coverage while maximizing the deployment efficiency, we study the partial sensing coverage problem and the corresponding node saving rate in 3D lattice WSNs. Two popular 3D deployment patterns including cube and triangular prism are considered. The partial sensing coverage and the note saving rate with respect to full sensing coverage are derived through mathematically modeling and theoretical analysis. Research results show that partial sensing coverage is of paramount significance to 3D lattice WSN design and implementation as a large amount of expensive 3D sensors can be saved by sacrificing a small amount of sensing coverage. For example, 38.43% and 22.14% sensors can be saved when providing 0.9864 and 0.9898 sensing coverage with respect to full sensing coverage in a cubic pattern and a triangular prism pattern based lattice WSN respectively. Computer-based simulations results validate the modeling and analysis.