Objective With continuous progress of energy measurement, some new technologies and equipments for online determination of natural gas calorific value and relative density have emerged in natural gas industry. The physical property method for calculation of natural gas Z-factor specified in GB/T 17747.3-2011 Natural gas-Calculation of compression factor-Part 3: Calculation using physical properties is needed for the application of the before mentioned new measurement method. However, physical property method is seldom used since it was published, evaluating its applicability and accuracy is important for the improvement and effective application of new technologies and equipments in the energy measurement of natural gas.

Methods In view of this, this paper assumes that the Z-factor calculated from detailed composition method in GB/T 17747.2-2011 Natural gas-Calculation of compression factor-Part 2: Calculation using molar-composition analysis gives the reference value, then theoretically compared the two algorithms based on 149 natural gases composition data from seven different gas sources and some simulated natural gases to exam the applicability of the physical property calculation method.

Results The calculation bias of Z-factors between detailed composition method and physical property method are within ±0.10% for most of natural gases and the accuracy of gross calorific value and relative density shall be within ±0.50% at least to meet the 0.30% accurate requirements of calculated Z-factor by GB/T 18603-2014 Technical requirements of measuring system for natural gas. The study also discovers that bias of the calculated Z-factors between two algorithms can reach to surprising 0.50% when the content of heavy hydrocarbons is relatively high to reach to the upper limit of GB/T 17747.3-2011.

Conclusions For most of the different sorts of commercial natural gases, the physical property calculation method is applicable. However, in some particular cases such as natural gas with equivalent C₂⁺ content higher than 10%, the bias of the calculated Z-factors between two algorithms can reach to 0.50%. It is suggested to theoretically analyze the distribution of Z-factor deviations calculated by two methods before the final decision on energy measurement method is made.