Electric power load in Brazil: view on the long-term forecasting models

Abstract Paper aims This paper aims to discuss how the energy load forecasts used by the System Operator and the main agents of the sector are made, and for what purposes, besides discuss the forecast deviations of the ONS’s and EPE’s models and their consequences and costs to the agents involved. Originality Fill a gap in the Brazilian literature when dealing with the Electric Power Load formally, and the consequences of forecast deviation used by the operator of system. Research method Simulations are carried out to estimate the cost of deviation, in order to assess the impact of the load forecast on the system. Main findings As a result, there is an urgent need to clarify forecasting methods and strategies; to avoid, by the sector agents, justifications that optimistic forecasts are necessary as security measures. Implications for theory and practice There are evidence that small improvements in forecasting models imply significant cost reductions.

研究目标：本文旨在探讨电网调度运营商（System Operator）与电力行业主要参与主体所采用的电力负荷预测方法及其应用场景，同时分析ONS与EPE的预测模型偏差，以及该偏差对相关参与主体造成的影响与成本。
研究创新点：本研究填补了巴西学术文献的空白——此前尚未有研究正式围绕电力负荷及电网调度运营商的预测偏差所产生的影响展开系统性探讨。
研究方法：本研究通过数值模拟估算偏差成本，以评估负荷预测对电力系统的影响。
主要研究结果：研究表明，亟需统一明确负荷预测方法与策略；同时电力行业主体应避免以“乐观预测是必要的安全保障措施”为由进行相关辩解。
理论与实践启示：有实证表明，对预测模型进行小幅优化即可带来显著的成本缩减。