In the Smart Grid (SG) network, the sensitive data is communicated over the network and it prone to attacks. In order to overcome these attacks, security algorithms like cryptography and steganography are used. Steganography algorithms provide imperceptibility by hiding the secret data in the cover image whereas cryptography algorithms scramble the data using a private key and scramble data gives attention to the attacker. In steganography, the Least Significant Bit (LSB) is the most preferred data hiding algorithm. In the LSB algorithm, the data is split into 1-bit chunks and hide in the cover image LSB bit. The LSB algorithm provides less variability with less embedding capacity. In this paper, we have designed an algorithm that provides better embedding capacity and lesser variability. Initially, the cover pixel read is read and its intermediate bits 5th and 6th are taken as a reference to hiding k-bits of data using the LSB algorithm. The benefit of taking the intermediate bits of the cover pixel as a reference is that no need to communicate reference bits information with the receiver. After that, the LSB bit of the stego image generated after data hiding is circularly rotated in order to reduce the variability. The experimental results were performed on the standard dataset images and various performance metrics calculated for it. Lastly, the proposed algorithm is compared with the existing algorithms and found that the proposed algorithm achieves randomness and better embedding capacity.