You should check carefully all your data and your measurements, particularly the true internal diameter of the complete pipe and the flow rate. Any very (very) small error propagates drastically over the friction factor. According to your computation (The Moody Diagram is basically a representation of the Colebrook-White equation), the internal roughness of the pipe in the Colebrook-White equation would be negative. I have no experience with petroleum, but with water, for instance, this occurrence does not look very likely.
Dear Sirs, Can we assume the flow to be streamlined? Or is there any possibility of 'Channel-like' flow due to pipe being not fully filled due to the presence of vapors?
Note: the pipeline is always pressurized at a min. 2 bar that means it is packed, but vapors may be playing some role here.
For the pressure and Reynolds number you mention, and if a gas or vapor is present, probably you may have a turbulent two-phase pipe flow. A two-phase flow produces a higher head loss than a single phase flow for the same superficial velocity: liquid flow rate /transversal section (the head loss can be measured in column of liquid, m). One possible approximation for the head loss computation is the use of the Colebrook-White Equation for the mixture (that is, for the total flow rate) and then you affect the result by the liquid fraction. You should confirm also the average liquid proprieties (density and viscosity) and measure the volumetric flow rates of vapor (if applicable) and liquid.
If you can get the total flow rate, you just need to measure the liquid flow rate (for example and if possible by a volumetric process, perhaps one of the most simple and direct methods) and so you can confirm if the gas is present or not and its value. This reference that you can see in my research gate page can give you more information with respect to your original question (in it, I used several techniques for measuring volumetric flow rates).Diogo, A. F., and Vilela, F. A. (2014). “Head losses and friction factors of steady turbulent flows in plastic pipes.” Urban Water J., 11(5), 414–425.