While it may well be possible to have calibrations which have a log or exponential relationship in a Calibration Line, the original problem was simply a case of mistyping the calculation into whatever instrument was being used to perform the calculation. The calibration range was 0 - 0.1 mg/L which yielded area responses in the range 0 - 8742401.
The unknown sample had an area of 1227456 which should yield a concentration within the calibration range (not 118.15 as the poster was calculating)
with the linear equation y= 9E+07x + 61347, 1227456-61347/9e+07=0.012956.
However, if you miscalculate and use 1227456-61347/9e+07=118.15