Limitations such as lack of detected stationary signal and slow signal recovery after detection currently affect graphene-based chemi-sensors operating at room temperature. In this work, we model the behavior of a sensor in a test chamber having limited volume and simulating the environmental conditions. From this model, we mathematically derive the calibration method for the sensor. The approach, focused on the time differential of the signal output, is tested on multi-layered graphene (MLG)-based sensors towards the chosen target gas (nitrogen dioxide) in the range from 0.12 to 1.32 ppm. MLG acting as sensing layer is synthesized by chemical vapor deposition. Our study paves the route for a wider applicability of the analysis to calibrate the class of devices affected by non-stationary and recovery issues.