Both gas and liquid pipelines commonly have small mid-line deliveries and injections with pressure and flow meters. Tuning an online model to use the pressure meters at these locations can be difficult, as the computed delivery/injection flow rate is very sensitive to a variety of different sources of error. This paper presents a novel method of computing the resistance between the branch pressure meter and the main line, which is a key component of the required tuning. Examples are shown from simulated gas and liquid pipelines and an attempt is made to apply the method to a real gas pipeline.


Online pipeline models use the available instrumentation to drive a simulation which tries to match the conditions in the real pipeline as closely as possible. The underlying differential equations only require (and indeed can only take advantage of) one pressure or one flow measurement at each point fluid enters or leaves the system, plus one temperature measurement each place fluid enters the system. This, however, is usually much less than the number of available meters: typically there are pressure, temperature, and flow meters at custody transfer entry and exit points, and there may be additional mid-line instruments as well.

So what can be done with these extra meters? If the pipeline model and configuration data were entirely perfect, the extra meters wouldn't be needed: if the model was using a flow meter as a boundary condition for its partial differential equations, then a pressure meter located at the same spot would just read the same pressure that the model solution gave for that point. In practice, however, there are all sorts of sources of inaccuracy in models that the extra meters can be used to improve:

  • unknown pipe roughness: the roughness of a pipe changes over time due to corrosion and may also be a bit different from the textbook value for steel pipe due to welds, etc.

  • similarly, the pipe might have bends or local diameter restrictions due to open valves and the like, which also provide unexpected resistance to flow

  • pressure and flow meters can drift, leading to constant offsets or multipliers of the true value

  • fluid properties may not be known perfectly

  • pressure and flow meters may not be located exactly where the schematics say they are

  • the thermal environment (ground temperature, soil type and especially soil moisture content) is almost impossible to know exactly

This content is only available via PDF.
You can access this article if you purchase or spend a download.