In electrical distribution system measurements, noise arises from both sensor-related distortions and grid disturbances. Accurate noise extraction is crucial for identifying and mitigating its impact on sensitive equipment. While existing studies primarily focus on improving frequency and harmonic estimation using methods such as the Discrete Fourier Transform (DFT), Fast Fourier Transform (FFT), and Interpolated Discrete Fourier Transform (IPDFT), their application to noise extraction in distribution systems remains limited. This research introduces the IPDFT-based approach for noise extraction and quantification. The voltage signal is segmented into smaller time frames, and a Hanning window is applied to minimize spectral leakage. IPDFT estimates the fundamental and harmonic components, which are used to reconstruct a signal containing only these components. The frequency spectrum of this reconstructed signal is then subtracted from the original spectrum, isolating the residual noise spectrum-representing the combined measurement and grid noise. The proposed method is applied to real voltage data from a medium-voltage distribution system. Analysis reveals that noise is primarily concentrated in higher frequency ranges, particularly around 10 kHz. Noise quantification is achieved by averaging the squared amplitudes within each segment and further averaging across all segments, providing a numerical representation of noise levels. The results indicate that noise levels remain relatively stable over time, with minimal variation in mean and standard deviation (SD).