978-1-4673-9134-4/16/$31.00 ©2016 IEEE Error Correction: a Proposal for a Standard Harold Kirkham 1 , Artis Riepnieks 1 , Eddy So 2 and Jim McBride 3 1 Pacific Northwest National Laboratory harold.kirkham@pnnl.gov 2 National Research Council of Canada 3 JMX Services, Inc Abstract — Some of the errors in transducers such as instrument transformers can be corrected as part of the digital processing for the measurement. The instrument transformer can be characterized in such a way that allows the Transducer Electronic Data Sheet of IEEE Std 1451 to transfer the information to the measurement system. A modification would allow the measurement system to perform a high-quality self- calibration whenever a transducer was replaced. That levies requirements on the characterization accuracy of the instrument transformer. Index Terms — Transducer Electronic Data Sheet, transducer error, self-calibration, error correction. I. INTRODUCTION Measurement connects the physical world to a conceptual model: an equation. In a moving coil instrument, the “equals sign” is manifested by the balance of the torque generated by the current in the coil against the torque of a return spring. In a digital instrument, the equation solving is more obvious. The equation being solved is the measurand. It was shown in [1] that the notion that measurement is equation-solving has many useful consequences. The simplifications and approximations of linguistic labeling are done away with: a concrete definition is readily found for measurands that have been problematical in the past—frequency, for example, when frequency is changing. The residuals can be used to calculate a metric that indicates the quality of the measurement.[2] In this paper, we use the notion that the measurand is a model, an equation, to examine high voltage measurement. II. MEASUREMENT FRAMEWORK The measurement framework, drawn as a block diagram, establishes the relationships between parts of the measurement process well known to metrologists. It can also be drawn to show the transducers used, as in Figure 1. In the figure, the solid arrows represent physical links, the open arrows conceptual ones. Figure 1 shows a measurement system such as a phasor measurement unit (PMU), in which the measuring instrument is made as accurate as required, and calibrated from its input terminals. In use, the measurement result will inevitably contain the artifacts added by the transduction system, typically using instrument transformers (IT). Applied to high voltage measurement, an isolation transformer, a CCVT or some active system based on field measurement may be involved. The point is that the primary quantity (the thing of concern to the application) is separated from the measuring instrument. The realized quantity is supposedly a scaled copy of the original, but the scaling will be imperfect, to a greater or lesser degree. III. CHANGING THE EQUATION The artifacts of the transducer that cause differences from perfect scaling are generally known as the cause of Type B uncertainties. In particular, the scale factor could be off- nominal, and there will be some frequency-response effects. These uncertainties are generally characterized during calibration tests. For high voltage and current measurements, instrument transformers (IT) are commonly used, with specified “accuracy class.” Both IEC and IEEE specify the relationship between ratio error and phase error graphically. The IEC standard [3] shows the boundary of the permitted region by rectangular boxes on a plane of ratio error vs phase error for each IT accuracy class without restrictions on the range of their load power factor. The IEEE standard [4] uses parallelogram boxes to describe their accuracy class, with the load power factor being limited within the range of 0.6 to 1.0. There are proposals to change the IEEE standard to the “square box” accuracy classes. It is proposed that at least some of these artifacts can be compensated for by changing the solution method in the measurement algorithm. Calibration data for the transducer can be added to the information available to the algorithm, and Fig. 1. Measurement framework, with transducer included • U.S. Government work not protected by U.S. copyright