Application of EVM to determine pipe internal surface conditions in complex natural gas lateral networks

Abstract

This paper describes the use of the Error-in-Variable Model (EVM), a type of parameter estimation techniques, for the determination of the internal pipe wall surface conditions (effective surface roughness) of pipe segments in complex natural gas pipeline networks and sub-lateral gathering networks. This technique uses SCADA data collected over a period of time such as pressures, temperatures and flow rates at boundary points of the network, i.e. at receipt and delivery points. These data are used as input to the EVM along with pertinent objective functions that are based on dimensional mass, momentum and energy conservation equations. The main focus of the technique is to extract reliable and accurate information with regard to the internal wall surface condition of pipe segments within the network that are showing evidence of fouling or under-performing from the flow efficiency perspective. The EVM is most suited for this problem as it accounts for uncertainties in the SCADA data from errors in the measurement instrumentations, the time averaging processing of the data and the inherent assumption of the quasi-steadiness of the flow rates. Insight into the implication of pipe segments arranged in series or in parallel within the network is highlighted. Guidance as to the relationship between the number of necessary distinct data sets required for a given number of parameters (surface roughness of pipe segments) to be estimated and the number of independent objective functions that can be written for a given network is provided. Finally, the EVM technique is demonstrated by an existing sub-lateral natural gas receipt network feeding into a main gas transmission pipeline (DN750), 21.1 km upstream of a compressor station.