The determination of mean residence time using statistical moments: is it correct?

The estimation of mean residence time using statistical moment theory is shown to lead to the true mean residence time only in the case of a concentration-time curve which is a single exponential. In other cases, including the commonly encountered double exponential, there is a systematic error present which can lead to the result being incorrect by as much as a factor of 2. For the case of instantaneous input, an alternative estimation procedure is suggested which gives an estimate much closer to the true mean residence time.