Area Variation observed

Maximum time we are getting substantial variation in area injection to injection in same vial for auto sampler method, for Ethanol content test. It's a area ratio calculation where ethanol is standard and n-propanol is internal standard. Though it's a internal standard method but area Variation is too much higher which is not acceptable, but same method we are running in other same model gc , in that case area Variation not observed, please explain the probable cause.