Computational fluid dynamics (CFD) analysis carried out for a leading oil and gas operator has highlighted mismeasurement problems worth more than $1m annually.
“TUV SUD NEL was asked to investigate a buckled orifice plate (OP) discovered during a routine audit of a shared pipeline,” says CFD Team Manager Marc Laing. “We were asked to undertake the necessary analysis to determine the flow measurement error resulting from this problem.”
“Our assessment revealed that the meter was over-reading by 5 to 10%,” Marc explains. “Given the flow rate in the system, this allowed us to provide an annual value for the mismeasurement, which has allowed the client to begin the process of claiming compensation from the owner of the pipeline.”
“At TUV SUD NEL, we are experts in this area, given our work on the development of ISO 5167, which covers the measurement of fluid flow by means of pressure differential devices such as orifice plates,” Marc adds. “Drawing on our experience, we knew that the best way of calculating the flow measurement error was to use CFD.”
To undertake this assessment, Marc and his team used cutting-edge laser scanning technology to scan the broken orifice plate and create a 3D model. This allowed them to accurately assess the degree to which the buckled plate was deformed and to capture the surface profile needed for the CFD analysis.
To ensure accuracy, an ideally installed OP was modelled using fluid properties calculated from TUV SUD NEL’s in-house physical property package (PPDS). The results of this modelling exercise were compared to the results in the appropriate standard (ISO 5167-2). It was found that the maximum deviation from the discharge coefficient in the standard was 0.2%, which showed excellent agreement (given the uncertainty in the standard is around 1%).
The models were then run for the ‘as found’ buckled OP installation. The relevant data were extracted to allow the comparison of the discharge coefficient in the ideal installation with the discharge coefficient in the ‘as found’ case. This revealed that the flow measurement errors were in the region of 5 to 10%.
In cases like this often there can be a lot of debate surrounding the methods used to calculate such an error given the potentially large financial exposure of the mis-measurement. In this case it was clearly demonstrated that the method being used was well within the required uncertainty for gas flow measurement (1%).
The meter was over-reading by 5 to 10% which, given the flowrate, meant that the mis-measurement was worth more than $1 million annually.
For more details, contact Marc Laing.