The harmonic levels in a distribution system are becoming more important as the use of non-linear loads and new technology is increasing. Monitoring of the harmonic levels is important in order to manage them. Direct measurement of harmonics at all locations is too costly, yet obtaining a system-wide view is important.

Harmonic State Estimation (HSE) is a cost-effective alternative to direct monitoring of the harmonics. To date, it has been applied mostly to transmission systems but in this work, it is applied to an actual distribution system using actual measurements. The aim of this research is to find out how the measured power quality data can be utilised and used for the efficient and accurate estimation of harmonics by using HSE technique, and installing equipment at optimal locations of power distribution networks.

HSE is performed based on a robust mathematical technique, Singular Value Decomposition (SVD). The SVD based HSE algorithm is first validated with the simulation results from the harmonic penetration study on the model of an actual distribution feeder using Open Distribution System Simulator (OpenDSS) and MATLAB. The estimated results and the measured values are compared and found to be extremely close to each other. The validated algorithm is then applied to the actual distribution system. Harmonics are measured at eight locations in one of the feeders of an actual distribution system. Using the measurements and the system model, HSE is performed. The estimates and the actual measurements are compared. Comparison is performed for several scenarios assuming different number of measuring equipment available for taking measurements.

The observability of the system is also checked. Locations of the measuring equipment, which can give maximum observability with the least errors in the estimated values, are found. Such locations are taken as the optimal locations for the placement of the measuring equipment. A power quality assessment tool has been developed to evaluate and visualise the power quality indices based on the measurements taken by the power quality monitors in Unison Networks Limited.

It has been established that harmonic state estimation using SVD method can be applied to the distribution systems to estimate the harmonics at unmonitored locations using few measurements and the system information. Further, it has been shown how the optimal location of the measuring equipment can be found using the observability of the system and the minimum error in the estimates. A new and simple method of adding new equation to the measurement equation in order to improve observability has also been applied for the non-triplen odd harmonics. This research has thus been successful in providing the experimental evidence for the application of HSE (based on SVD) to the distribution systems for the first time, finding optimal location for the placement of measuring equipment to monitor harmonics, and developing Power Quality Assessment and Visualization Tool for the Power Quality Management System in a distribution company.