Topics In Demand
Notification
New

No notification found.

Energy Forecasting
Energy Forecasting

November 3, 2020

6

0

 

Energy forecasting is a technique used by power companies to predict the power or energy needed to balance the demand at all times. It is essential to provide decision making support for resources planning and operations of power utilities.  The main aim is to make proper capacity utilization, accurate demand and supply planning, lowering of the penalty imposed by the Regulator. In a way, it is beneficial for both buyers as well as sellers, creating a win-win solution for all.in both the short term and long term. This includes forecasting demand and price of electricity, fossil fuels such as natural gas, oil and coal and renewable energy sources such as hydro, wind and solar. In the recent past, the need of  forecasting demand for an electric utility has become a much-discussed issue. In the earlier times, the power grid used to function in isolation, so the regions falling short of power used to be imposed with prolonged power cuts, while the regions having surpluses could do nothing to avert power spillage. So, in such an isolated scenario,  setting up power plants in the respective regions used to suffice the demand pattern. But with the modernization of the technology, integrated functioning of the grid and an exponential growth in the electricity demand pattern, the need arises for an enhanced level of energy forecast. 

The most popular types of energy forecasting are:

  • Load Forecasting
  • Electricity Price  Forecasting
  • Wind power forecasting
  • Solar power forecasting

In load forecasting, the “load” means electricity demand (in kW/MW) or energy (in kWh/MWh). Since the magnitude of power and energy is the same for hourly data, no distinction is made between demand and energy. Load forecasting or demand forecasting is a process where historical trends of demand consumption vis-à-vis weather profile analyses assist in  estimating   electricity demand at consumer end . Temperature, Humidity, Weather Conditions and Day Type (Weekdays, Weekends, festivals etc.) are considered as they are significant factors impacting the effectiveness of an accurate Day Ahead Forecast. The algorithm makes use of Artificial Intelligence (AI) and Machine learning based formulation for Energy Forecast combined with Statistical Techniques. The model is capable of predicting the next 24-hour demand of the state using the Weather parameters of previous and forecasted day(s) and energy data of previous day(s) to visualize and capture the trend. Business assumptions like turnover, profit margins, cash flow, capital expenditure, risk assessment and mitigation plans, capacity planning are some of the factors which depend on Demand Forecasting.

Electricity price forecasting works by predicting the spot and forward prices in wholesale electricity markets.

Forecasting of wind power generation occurs at different time scales. That means it can vary from milliseconds up to a few minutes. It provides an estimate of the expected generation of a few of multiple wind turbines (with units kW or MW depending on the wind farm nominal capacity).The forecasts can also be expressed in terms of energy, by integrating power production over each time interval.

Solar power generation forecasting is categorized into two parts. First is forecasting of solar irradiance  and/or any other meteorological variable and second is estimating the amount of energy that a solar power plant will produce with the estimations provided by the respective  meteorological resource. This type of forecasting involves: 

  • The atmospheric  conditions
  • The cloud movements and scattering processes 
  • The characteristics of a solar energy plant which utilizes the Sun’s energy to generate  solar power
  • Solar photovoltaic systems which transform solar energy into electric power.

Some of the common solar forecasting methods are stochastic learning methods, gradient boosting, random forest, local and remote sensing methods, and hybrid methods.

Forecasting horizons:

Short-term forecasting refers to forecasting from a few minutes up to a few days ahead. It means immediate and thus requires immediate forecasting techniques.

Medium-term forecasting refers to forecasting from a few days to a few months ahead. It is generally preferred for balance sheet calculations, risk management and derivatives pricing.

Long-term Forecasting refers to times measured in months, quarters or even years. It focuses on investment profitability analysis and long term planning such as determining the future sites or fuel sources of power plants.

.Kreate provides RED – a highly accurate machine learning and statistics based Renewable Energy generation and Demand forecasting tool. It is used to meet the needs of generators, distribution companies, state and regional load dispatch centers, as well as assists in complying to the regulatory norms of the Hon’ble commissions. Except for the days having unpredictable breakdowns due to issues such as load crash/telemetry issue/force majeure conditions, this model gives accuracy more than 97%. To achieve this high level of performance, blending of good quality weather with real time energy data are pre-requisites. 

It has features such as:

  •  Real-Time Analytical Dashboard and Mobile application for monitoring of Solar/ Wind Plant generation data, demand data, comparing actual with forecast and monitoring Deviation Settlement Mechanism (DSM) charges.
  • Monthly MIS reports on forecast vs actual generation, DSM charges summary and any other MIS as per the requirement.
  • SMS/Email alerts to the concerned officials.
  • Compatible with almost all Browsers.
  • Available on cloud and for in-house installation.
  • Our reporting format meets all CERC/SERC compliance requirements.
  • Robust QA/QC process is followed to ensure the best forecasting result.
  • Fast customer set-up time.

CHALLENGES:

Weather Data Challenge: At times, the weather forecast data is quite deviated from the actual weather behavior. In addition, historical weather is available on a 3-hour interval and forecast weather is available on an hourly basis. So in order to make use of this hourly data with the 15-min block-wise demand and generation data, certain mathematical techniques are employed. These techniques, despite being dependable, are responsible at some level to induce error.

Wide Weather Variations Geographically: Across any state, wide weather variations are noticed. So if disturbed weather is not observed in the areas contributing to major drawl of power, there would be quite less/no modification in the demand pattern. On the contrary, the area drawing major share of power has a disturbed weather and the rest of the state has a normal weather profile, there would be a great dip in demand pattern.

Another point of concern being weather departments that issue forecast & actual weather data are not available in 15 min block wise manner (actual data available in intervals of 3hrs for most of the geographical locations).

 

Related to Real-Time Demand and Generation Data:

  • Unpredicted breakdowns due to thunderstorms and rain: In most of the states, whenever turbulent weather conditions occur, the unaccounted breakdowns occur. In addition, the lightning and thunderstorms induce unwanted spikes/dips in the telemetry recording mechanism.
  • As the solar and wind plants are located at the outskirts of any city, so due to poor network connectivity, the actual generation data is received with a lag or 1-1.30 hrs, which reduces the quality of generation.

 


That the contents of third-party articles/blogs published here on the website, and the interpretation of all information in the article/blogs such as data, maps, numbers, opinions etc. displayed in the article/blogs and views or the opinions expressed within the content are solely of the author's; and do not reflect the opinions and beliefs of NASSCOM or its affiliates in any manner. NASSCOM does not take any liability w.r.t. content in any manner and will not be liable in any manner whatsoever for any kind of liability arising out of any act, error or omission. The contents of third-party article/blogs published, are provided solely as convenience; and the presence of these articles/blogs should not, under any circumstances, be considered as an endorsement of the contents by NASSCOM in any manner; and if you chose to access these articles/blogs , you do so at your own risk.


adil

© Copyright nasscom. All Rights Reserved.