A note on estimating quantiles of exponential populations