I have EIS tests with and without inhibitor, but attached Nyquist plot is not starting at Zero point in in real Z values? the inhibitor is very strong and it should achieve wide diameter semicircle.
Depending on your specific setup you will have a resistance in series with your interface that has contributions from solution resistivity, connectors and wires. In a Bode plot you will typically see this in the high frequency end of the spectrum as a plateau. Since this is a series resistance (not frequency dependent) using a different frequency range will not help here.
Johannes is correct; the high-frequency limit (at zero reactance) is the series resistance. The large difference in series resistance between the two measurements indicates some connection problem.
The expected solution resistance can be estimated readily from the size and shape of the working electrode, and the conductivity of the electrolyte (which seems to be 3% NaCl from the caption on the figure). This is a highly conductive electrolyte, so the solution resistance should be small.
I agree with Christiaan and Johannes. Higher frequency range won't help. One sample was badly contacted and has ≈ 50ohm contact resistance (also normalized by the area), which does not disturb the whole analysis. It seems that you have at least two (RC) time-constants, roughly indicated by the semicircles in blue. The spectrum with red points is more complicated. It does not seem to be stationary and the larger red semicircle is "growing" during the measurement: As you decrease the frequency the measuring times grows considerably and the resistance associated with the second capacitive loop is growing (full to dotted line). In this case, try to wait longer before you start the measurement. A straight part with 22,5º degrees (red line) could indicate the existence of partial diffusion control. Try to use the same scale in both axes (ohms.cm2/scale cm).