Challenges in Demand Forecasting Raj Protim Kundu ERLDC, POSOCO - PowerPoint PPT Presentation

Challenges in Demand Forecasting Raj Protim Kundu ERLDC, POSOCO

Need for forecasting  IEGC mandates it - “ 5.3 (C) Each SLDC shall develop methodologies/mechanisms for daily/ weekly/monthly/yearly demand estimation (MW, MVAr and MWh) for operational purposes. Based on this demand estimate and the estimated availability from different sources, SLDC shall plan demand management measures like load shedding, power cuts, etc … . ”  For better planning - Electricity can not be stored in large quantum in economical way. - If area wise demand can be forecasted well in advance , uninterrupted, reliable power can be delivered - Increase in renewable energy will increase more uncertainties in supply side also January 22, 2019 (c) POSOCO 3

Types of load forecasting January 22, 2019 (c) POSOCO 4

Accuracy and usages of different types of load forecasting January 22, 2019 (c) POSOCO 5

Eastern Region Demand Variations for 2017-18 January 22, 2019 (c) POSOCO 6

Input data sources for STLF Real time Historical Load & Weather weather data data base Forecast Measured load STLF Information EMS display January 22, 2019 (c) POSOCO 7

Load Forecasting Model Development January 22, 2019 (c) POSOCO 8

Challenges in forecasting  Selection of proper forecasting models - What influencing factors to be considered. - Operational experiences are important  Quality of input data - Unconstraint demand data are required  Selection of forecasted area - Demand of large control area dependent on large no of parameter - Demand of small control area dependent on connectivity with rest of the grid  Sudden contingencies - Loss of important generating units or transmission/distribution elements  Sudden weather changes - Storm, Heat waves, Cold waves, Humidity changes, Fog January 22, 2019 (c) POSOCO 9

Load crash due to Titli on October 11, 2018 (Thursday) EASTERN REGIONAL DEMAND DATE : 11 Oct 2018 Thursday Actual Demand Time DMD 10-10-18 18:51 19478 Reg.Max (MW) 11-10-18 Reg.Min (MW) 6:34 15324 22000 20000 ---> W) --- 18000 mand (MW) 16000 14000 Dema 12000 10000 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Hours of the Day Around 2500-3000 MW reduction in regional demand was observed in the early morning hours on 11-10-2018. January 22, 2019 (c) POSOCO 10

Load crash due to Titli on October 11, 2018 (Thursday) Around 1000 MW demand reduction was observed for the first few hours on 11-10- 2018 January 22, 2019 (c) POSOCO 11

Resources for thoughts  U. K. Verma, S Banerjee, R P Kundu, Comparison of different forecasting models used for short term load forecasting, CBIP water and energy international journal May 2016  V. K. Srivastava, S Mishra, V Pandey, S S Raghuwansi and A Ahmed, Load and RE Forecasting- Utilization and Impact on System Operation, CIGRE – AORC Technical Meeting 2018  POSOCO, Report on Electricity Load Factor in Indian Power System, 2016  E A. Feinberg, D. Genethliou, Load Forecasting, Applied Mathematics for Restructured Electric Power Systems: Optimization, Control, and Computational Intelligence, pp. 269-285, Spinger, 2005.  A. Meyler, G. KENNY, T. QUINN, Forecasting Irish Inflation using ARIMA Models, Research and Publications Department, Central Bank of Ireland, , Vol. 3/RT/98, December, 1998.  R. J. Hyndman and Y. Khandakar, Automatic Time Series Forecasting: The forecast Package for R, Journal of Statistical Software, Volume 27, Issue 3, July 2008. January 22, 2019 (c) POSOCO 12

Conclusion “The more you sweat in peace, the less you bleed in war.” - Norman Schwarzkopf rajprotim@posoco.in 9903329591 Thank You !! 1/22/2019 (c) POSOCO 13