Nowadays, information and communication technology (ICT) has become more and more energy conscious. In this paper, we focus on peer-to-peer systems which contribute a major fraction of the Internet traffic. This paper proposes analytical models of energy consumption in P2P system. The model considers content pollution, the most common attack in P2P system, which has received little attention in previous work on green P2P. The analysis of the models shows that the popular sleep method in green computing potentially affects peer-to-peer performance. When the online time of clean copy holders is over cut, the system collapses. To find the balance between energy saving and system maintenance, the concept energy effectiveness is introduced. An approach for controlling energy consumption while keeping the system stable is suggested. We show that the whole system can be benefited if some warm-hearted and smart peers are willing to spend a little extra cost on energy, when most peers over cut their power on time. This approach can perfectly complement the popular sleep methods in green computing.