A Comparative Study of Iterative and Non-Iterative Load-Flow Methods: A Case of Newton-Raphson and Holomorphic Embedding Approaches

Abstract

Given the crucial role load flow analysis plays in the planning and operation of power systems, there is a growing requirement to produce load flow  methods that would perform with unerring accuracy and also be free from the convergence concerns associated with the classical iterative  methods. The Holomorphic Embedding Load-flow Method (HELM) is one of such attempts. It is non-iterative and could produce a solution when it is  available and indicate when there is no solution to signal an abnormality, such as a voltage collapse. According to literature, the Newton-Raphson  load flow method (NRLM) is the most widely utilized iterative method due to its remarkable qualities, therefore it is the preferred choice here for  comparison. The objective of this paper is to weigh up the merits of HELM over the NRLM based on information obtained from actual applications  and since HELM is not found as one of the methods previously applied for the analysis of the Nigerian network, the behaviour of the system with  HELM is assessed. Analyses of particular systems were done at least three times and the average time for each was computed. HELM was found to  be faster by a wide margin in all the cases. For example, HELM was 95.7% faster than the NRLM for the 4-bus system and 80.6% faster when applied  to the Nigerian 330kV transmission network. This validates a major advantage of HELM over iterative solutions. However, the node voltages of  Nigerian system were not as close as results for standard IEEE systems which can be attributed to the condition of the network which HELM, as one  of its advantages is believed to expose. Secondly, though the special software produced for HELM analyses might be the best for it, an attempt was  made to adapt to a MATLAB program since the software is easily accessible. More work is required in the programming to successfully analyse large  and ill-conditioned systems.