A real data set for a 116-node power transmission system

Published: 04-07-2018| Version 1 | DOI: 10.17632/dv3vjnwwf9.1
Contributors:
Meltem Peker,
Ayse Kocaman,
Bahar Kara

Description

This data is used in “A Two-Stage Stochastic Programming Approach for Reliability Constrained Power System Expansion Planning” article by Peker, M., Kocaman, A. S., and Kara, B. Y. (2018). (Peker, M., Kocaman, A. S., & Kara, B. Y. (2018). A two-stage stochastic programming approach for reliability constrained power system expansion planning. International Journal of Electrical Power & Energy Systems, 103, 458-469.) We present an open data set for the Turkish 400-kV transmission network that includes 116 nodes and 190 transmission lines, which is referred to as "Reduced Turkish Power System". Coordinates, hourly loads and installed capacities of nodes, and characteristics of existing transmission lines for this system are presented. Suitable locations for expansion plans are also provided. Turkish power system has 970 counties, which are considered as demand points, and 997 substations and 1746 transmission lines which are comprised of 400-kV, 220-kV, 154-kV and 66-kV voltage levels. We focus on the backbone of the Turkish transmission system that consists of 400-kV substations and 400-kV transmission lines. We observe that 116 counties out of 970 have at least one 400-kV substation. We implement the power-island model explained in (Singh et al., 2014) to aggregate the hourly demand profiles and installed capacities of all counties to these 116 substation nodes. Counties, power-island assignments, substation nodes, and transmission lines between the 400-kV substation nodes are presented in Reduced Turkish Power System.png. The data is presented in tables in Microsoft Excel (.xlsl) and shapefiles are also provided to regenerate the figure. Data are obtained from online databases, which are publicly available, and the references are also presented.

Files

Steps to reproduce

The raw data utilized to generate this dataset is also presented. Since the hourly demand profiles of counties are not directly available, they are estimated based on their populations. The hourly distribution of demand for 21 distribution companies (DC) in Turkey are provided. These hourly values, however, are normalized with the amounts observed on the first Monday of each month. Therefore, to obtain the actual values, another dataset that provides the monthly demand values of 81 cities in Turkey is used. Given that each city is served by only one DC, the hourly demand values of each city in Turkey are obtained. Finally, to obtain the hourly demand profiles of 970 counties in 81 cities, hourly demand amounts observed in the cities are scaled with the populations of the counties.