Projects: Projects for Investigator |
||
Reference Number | NIA_NGSO0020 | |
Title | Short-term System Inertia Forecast | |
Status | Completed | |
Energy Categories | Other Power and Storage Technologies(Electricity transmission and distribution) 100%; | |
Research Types | Applied Research and Development 100% | |
Science and Technology Fields | PHYSICAL SCIENCES AND MATHEMATICS (Computer Science and Informatics) 70%; ENGINEERING AND TECHNOLOGY (Electrical and Electronic Engineering) 10%; ENVIRONMENTAL SCIENCES (Earth Systems and Environmental Sciences) 20%; |
|
UKERC Cross Cutting Characterisation | Not Cross-cutting 100% | |
Principal Investigator |
Project Contact No email address given National Grid plc |
|
Award Type | Network Innovation Allowance | |
Funding Source | Ofgem | |
Start Date | 01 March 2019 | |
End Date | 01 September 2020 | |
Duration | ENA months | |
Total Grant Value | £300,000 | |
Industrial Sectors | Power | |
Region | London | |
Programme | Network Innovation Allowance | |
Investigators | Principal Investigator | Project Contact , National Grid plc (100.000%) |
Industrial Collaborator | Project Contact , National Grid plc (0.000%) |
|
Web Site | https://smarter.energynetworks.org/projects/NIA_NGSO0020 |
|
Objectives | This innovation project involves, for the first time, investigation of the feasibility of a data-driven approach to provide multi-time resolution inertia forecasts with high accuracy. The project will involve the following activities and methodological approach: Collect historical data for system conditions and estimate inertia levels. The most critical data is the historic or real time total system inertia measurements. Subject to availability, other data may be used to improve the forecasting accuracy of the system inertia, which includes weather data (such as temperature, wind), system condition data (such as National Grid ESOs Integrated Energy Management System (IEMS), PMU, BMU and so on) and forecasting data (such as demand forecast and renewable energy forecasts). Apply data-driven approach (e.g LASSO) to identify the most relevant features (temperature, hour of the day etc.) related to the frequency (PMU data) and inertia (ROCOF data), contributions from synchronous generation (BMU data), demand side and distributed generation. Develop Machine Learning-based predictive models (e.g., generalized linear models, deep learning, etc.) for multi-resolution point and probabilistic inertia forecasts in a rolling basis. Apply advanced risk-constrained system scheduling model and frequency response market-clearing models to quantify the impacts and benefits for accurate inertia forecasts. Due to the complicated dependency structure and lack of detailed measurements in the demand side, there is very limited understanding on the inertia contribution from demand side and embedded units. Traditional physical modelling based approach is hence not applicable in this case. Therefore, in the context of this project, the proposed activities and methodology will develop advanced Machine Learning-based predictive models (e.g., random forest, deep learning, etc.) for multi-temporal inertia forecasts in a rolling-basis. Based on available data, this will also try to identify the most relevant features (temperature, hour of the day, demand forecast etc.) related to the inertia contribution from demand side by applying automatic feature selection methods. This project aims to provide a proof of concept tool for an accurate day-ahead and intra-day inertia forecast with multi-time resolution, that can be potentially used to support the day-ahead frequency response procurement and the real-time system operation. | |
Abstract | This innovation project involves, for the first time, investigation of the feasibility of a data-driven approach to provide multi-time resolution inertia forecasts with high accuracy. | |
Publications | (none) |
|
Final Report | (none) |
|
Added to Database | 09/11/22 |