PR-015-09605-R01 Extended Low Flow Range Metering

Natural gas meters are often used to measure flows below their minimum design flow rate. This can occur because of inaccurate flow projections, widely varying flow rates in the line, a lack of personnel available to change orifice plates, and other causes. The use of meters outside their design ranges can result in significant measurement errors. The objectives of this project were to examine parameters that contribute to measurement error at flow rates below 10% of a meters capacity, determine the expected range of error at these flow rates, and establish methods to reduce measurement error in this range. The project began with a literature search of prior studies of orifice, turbine, and ultrasonic meters for background information on their performance in low flows. Two conditions affecting multiple meter types were identified for study. First, temperature measurement errors in low flows can influence the accuracy of all three meter types, though the effect of a given temperature error can differ among the meter types. Second, thermally stratified flows at low flow rates are known to cause measurement errors in ultrasonic meters that cannot compensate for the resulting flow profiles, and the literature suggested that these flows could also affect orifice plates and turbine meters. Several possible ways to improve temperature measurements in low flows were also identified for further study. Next, an analytical study focused on potential errors due to inaccurate temperature measurements. Numerical tools were used to model a pipeline with different thermowell and RTD geometries. The goals were to estimate temperature measurement errors under different low-flow conditions, and to identify approaches to minimize temperature and flow rate errors. Thermal conduction from the pipe wall to the thermowell caused the largest predicted bias in measured temperature, while stratified temperatures in the flow caused relatively little temperature bias. Thermally isolating the thermowell from the pipe wall, or using a bare RTD, can minimize temperature bias, but are not usually practical approaches. Insulation of the meter run and the use of a finned thermowell design were practical methods predicted to potentially improve measurement accuracy, and were chosen for testing.