ISO 3924 Petroleum products - Determination of boiling range distribution - Gas chromatography method
7 Sampling
Unless otherwise specified, samples shall be taken by the procedures in accordance with ISO 3170 or ISO 3171.

8 Preparation of apparatus
8.1 Column preparation
8.1.1 General
Any satisfactory method that produces a column meeting the requirements of 6.2 may be used. The column shall be conditioned at the maximum operating temperature to reduce baseline shifts due to bleeding of the column substrate.

8.1.2 Packed columns
An acceptable method of column conditioning that has been found effective for columns with an initial loading of 10 % liquid phase consists of purging the column with carrier gas at the normal flow rate while holding the column at the maximum operating temperature for 12 h to 16 h.

8.1.3 Capillary columns
Capillary columns may be conditioned using the following procedure.
a) Install the column following the manufacturer's instructions. Set the column and detector gas flows. Ensure that the system is leak-free.
b) Allow the system to purge with carrier gas at ambient temperature for at least 30 min. Then increase the oven temperature by approximately 5 °C/min to 10 °C/min to the final operating temperature and hold for approximately 30 min.
c) Cycle the chromatograph through its temperature programme several times until a stable baseline is obtained.

NOTE Capillary columns with cross-linked and bonded phases are available from many manufacturers and are usually preconditioned. These columns have much lower column bleed than packed columns.

8.2 Chromatograph
Place the chromatograph in service in accordance with the manufacturer's instructions. Typical operating conditions are shown in Tables 2 and 3.

If a flame ionization detector is used, the deposits formed in the detector from combustion of the silicone decomposition products shall be removed regularly, as they change the response characteristics of the detector.

NOTE Without any instrumental adaptation, it is possible to decrease analysis time, as described in Annex B.

8.3 Column resolution
Analyse the calibration mixture under the same conditions as those used for the samples. Using the procedure illustrated in Figure 1, calculate the resolution, R, from the time between the C16 and C18 alkane peaks at the peak maxima, t1 and t2, and the widths, y1 and y2, of the peaks at half height, as given by Equation (1):
R = 2(t2 - t1)/1.699(y1 + y2)
t1 is the retention time, expressed in seconds, for the C16 peak maximum;
t2 is the retention time, expressed in seconds, for the C18 peak maximum;
y1 is the width, expressed in seconds, at half height of the C16 peak;
y2 is the width, expressed in seconds, at half height of the C18 peak.

The resolution, R, obtained from Equation (1), shall be at least 3.

8.4 Detector response check
This method assumes that the detector response to petroleum hydrocarbons is proportional to the mass of individual components. This shall be verified when the system is put into service and whenever any changes are made to the system or operational parameters. Analyse the calibration mixture using the same conditions as those used for the samples. Calculate the response factor, Fn, for each alkane relative to decane using Equation (2):
Fn = (mn/An)/(m10/A10)
mn is the mass of the alkane in the mixture;
An is the peak area of the alkane;
m10 is the mass of decane in the mixture;
A10 is the area of decane.

The relative response factor, Fn, of each alkane shall not deviate from 1.0 by more than +/- 0.1.

8.5 Peak skewness
Determine the peak skewness (the ratio A/B) of the largest peak in the calibration mixture as shown in Figure 2.

The peak skewness shall be not less than 0.5 and not more than 2.0. If peak skewness is outside these parameters, reanalyse the calibration mixture using a smaller sample size or a more dilute solution, if necessary, to avoid peak distortion.

NOTE Skewness is often an indication of overloading the sample capacity of the column, which results in a displacement of the peak apex relative to non-overloaded peaks. Distortion in retention-time measurement, and hence errors in boiling point determination, are likely if column overloading occurs. The column liquid-phase loading has a direct bearing on the acceptable sample size.