go to top scroll for more

Projects: Projects for Investigator
Reference Number	NIA_NGET0183
Title	Solar PV Forecasting Phase 2
Status	Completed
Energy Categories	Renewable Energy Sources(Solar Energy, Photovoltaics) 75%; Other Power and Storage Technologies(Electricity transmission and distribution) 25%;
Research Types	Applied Research and Development 100%
Science and Technology Fields	PHYSICAL SCIENCES AND MATHEMATICS (Applied Mathematics) 50%; ENGINEERING AND TECHNOLOGY (Electrical and Electronic Engineering) 25%; ENVIRONMENTAL SCIENCES (Earth Systems and Environmental Sciences) 25%;
UKERC Cross Cutting Characterisation	Not Cross-cutting 100%
Principal Investigator	Project Contact No email address given National Grid Electricity Transmission
Award Type	Network Innovation Allowance
Funding Source	Ofgem
Start Date	01 July 2016
End Date	01 July 2018
Duration	24 months
Total Grant Value	£300,000
Industrial Sectors	Power
Region	London
Programme	Network Innovation Allowance
Investigators	Principal Investigator	Project Contact , National Grid Electricity Transmission (100.000%)
Web Site	http://www.smarternetworks.org/project/NIA_NGET0183
Objectives	The purpose of this project is To derive datasets and specific knowledge of characteristics of solar PV generation in terms of variability, ramping and persistence, and the joint characteristics of how the solar resource interacts with the wind resource, that will inform planning decisions and procurement of balancing services within NGET. To develop new models for converting solar irradiance into generated solar PV power. This will improve the accuracy of both solar generation and transmission demand forecasts and hence expenditure on constraint management and reserves. To improve short term solar generation forecasts The objectives of this project can be considered in the following specific areas. Variability of solar irradiance and its correlations with other meteorological variables As part of work package 1, NGET will be provided with a 34-year time-series of synthetic GB-aggregated solar irradiance, wind speed and temperature and an accompanying report. This will provide details of the variability in solar irradiance and frequency and duration of extreme and unusual events over a range of time scales; including decadal, annual and seasonal. In addition it will highlight the frequency and magnitude of ramping events. This will allow current events to be placed in historical context and enable a better insight of the potential impacts of solar PV generation on the power system. The data also could be used in combination with data provided in a previous NIA project (NIA_NGET0016: UK-wide wind power resource: Extremes and variability), to consider extreme events associated with combined wind and solar generation. This information could be harnessed for operational planning or for investigating future energy scenarios. Develop improved transfer model for solar power generation; determine characteristics of solar generation and interaction with wind generation and demand variability The objective of work package 2 is to provide NGET with improved models for converting solar irradiance into solar generation power. This will improve the solar generation forecasts and the transmission demand forecasts. Using the results from Work Package 1 a joint time series of solar load factors, wind load factors and demand at regional and aggregated GB levels will be provided. A report will be presented detailing statistical characteristics, both individually and jointly of the renewable resources, focussing on variability, ramping and persistence. A library of extreme or difficult events on the transmission network will be included. Assess the possible improvements in short term solar radiation forecasts On completion of work package 3, a report will be presented to NGET detailing the use of remote sensing techniques and statistical techniques driven by near-real time solar generation data in improving short term solar forecasts. This project will provide knowledge that; (1) enables a better insight of the potential impacts of solar PV generation on the power system, at planning time scales and also for assessing tools needed within balancing services (2) improves medium and short term solar generation forecasts and hence demand forecasts. This activity will contribute to cost reduction in an area of high strategic significance. However, the full attribution of constraint and balancing costs is highly complex, rendering any estimate of direct cost benefit misleading at best. Success of this project will be evidenced by availabilityof solar generation information within the operational planning process. This information can be aligned to three main objectives; (1) Greater awareness and use across NGET of impact of solar generation variability and its interaction with wind generation and demand variability for use in network and commercial planning (WP 1 and 2) (2) Improved medium term solar generation forecasting for use in operational and commercial decision making (WP 2) (3) Improvements in short term solar radiation forecasts to improve operational decision making (WP 3).
Abstract	A recent, dramatic increase in installed photovoltaic generation is now impacting the electricity demand profile. This influence has been challenging to predict and is currently leading to significant demand forecast errors. The total solar capacity in Great Britain is now in excess of 9. 3 GW, and is forecast to rise to 15. 7 GW by 2020. Owing to the size of the individual installations (the largest solar farm in the UK is just 48 MW) all of this capacity is embedded within the distribution networks. Currently National Grid has a simplified transfer model to convert solar irradiance into generated power. This model has not been tested against actual solar generation data because of the lack of available data. National Grid has identified a number of opportunities to improve methods of forecasting solar PV generation in order to reduce electricity demand forecast errors. This project compliments the work being done under the following NIA projects that have recently been started over the past six months: NIA_NGET0139 - Solar PV Monitoring Phase 1. NIA_NGET0170 - Solar PV Monitoring Phase 2. NIA_NGET0177 - Solar PV Forecasting Phase 1. There is also a need to consider effects of the variability of solar generation, and how it interacts with other variable generation, particularly wind, at higher penetration levels. This impacts both on design of network in planning time scales and on economic tools to manage the variability. The scope and methods for the complimentary solar monitoring and forecasting projects identified above are summarised as follows: NIA_NGET0139 - Solar PV Monitoring Phase 1 and NIA_NGET0170 - Solar PV Monitoring Phase 2 are focused on developing the capability for the Electricity National Control Centre to be able to interpolate and accurately model solar output using a sample of real data from the many hundreds of thousands of solar PV installations already operating and those that may be connected to the network in the future. NIA_NGET0177 - Solar PV Forecasting Phase 1 is being undertaken with the Met Office to test and stretch the current state of the art capabilities cloud cover forecasting (the largest single factor affecting of solar pv forecast error) at a scale that is meaningful for forecasting regional output from solar PV installations. This project builds on the work done previously by National Grid in the context of wind forecasting by re-analysing historic weather data to model how a power system with today’s capacity of solar PV would have behaved at different times of the past 3 decades to provide a context to the potential frequency of extreme events. The project partner has developed a range of tools and approaches in previous wind power investigations. Archived data from Numerical Weather Prediction (NWP) models provides a history of atmospheric behaviour and weather forecast predictions. Satellite based Re-analysis datasets draw on NWP models to provide a 34 year, large scale representation of the atmosphere, from which atmospheric properties over Great Britain can be derived. This project will leverage the expertise and approaches from the earlier work on wind generation. The Re-analysis datasets will be used to construct a 34 year, coherent representation of solar irradiance, wind speed and temperature over a GB grid. This data can be used to derive distributions of solar variability, solar ramping events and combined solar/wind events. These distributions will be found at different time scales, from hourly to daily, seasonally or yearly, and different geographical scales, from regional to aggregate GB. The precise finest time or spatial scale will be determined during the data-exploratory phase of the project, but is expected to be: temporally - hourly; and spatially - Grid Supply Point (GSP) group level. This project will construct a validated solar irradiance to electrical power model suitable for being used at national or regional level. This will improve the current crude model used by National Grid, but will also avoid the unnecessarily detailed modelling that is done at solar farm level. This will both improve operational solar generation forecasts, and hence electricity demand forecasts; but also allow the construction of joint time series of demand, solar generation and wind generation which are vital for planning network requirements and balancing services requirements. UoR has expertise in using satellite and radar data for improving short term forecasts of precipitation in between the 6 hourly Met Office weather model runs. This project will leverage this expertise to use satellite and radar information to improve solar forecasts from 1 hour to 6 hour time horizon, improving operational visibility of potential changes in the highly volatile solar generation. The activity will be structured in three work packages, aligned to the objectives described below. WP1. Determine characteristics of variability of solar irradiance and its correlations with other meteorological variables In order to understand the characteristics of solar PV generation, particularly high impact low probability events, we need to understand variability of atmospheric solar irradiance, and how it interacts with other meteorological variables, particularly wind speed and temperature. Observations for solar irradiation are not available at many of the Met Office weather stations that now exist, or have only recently been recorded. In addition, there are a limited number of Met Office weather stations located close to the solar PV installations. This work package will use state of the art meteorological models and datasets to construct a joint time series of solar irradiance, wind speed and temperature. It will use this time series to derive statistical characteristics, particularly variability, ramping and persistence of the renewable resource at different timescales Use state-of-the-art meteorological datasets to generate a long term joint time-series (34 years) of solar irradiance, wind speed, temperature at finest achievable resolution, but no less than 3-hourly and regionally. Validate the time series against observed data where available to assess to what extent the time series is capturing the full range of extremes. Quantify the magnitude and frequency of a range solar irradiance events, particularly variability, ramping and persistence, and their joint occurrence with other rare meteorological events DELIVERABLE: NGET will be provided with a 34-year joint time-series of synthetic regional and GB solar irradiation, wind speed and temperature and an accompanying report. This will provide details of the variability in solar irradiance over a range of time and spatial scales; including decadal, annual, seasonal and daily, as well as the correlation of wind and solar variables. WP2. Develop an appropriate transfer model from solar irradiance to solar power generation, and apply it to thetime series from WP1 to determine characteristics of solar PV generation This work package will develop the transfer model that will be at the heart of National Grid’s solar power forecasting, improving the current model in use. It will apply this model to the synthetic time series in WP1 to assess the characteristics of solar PV generation variability, and its correlations to wind generation and to synthetic demand time series. Develop improved transfer models to convert solar irradiance into generated solar PV power and validate these against metered output solar power, improving medium term (6 hours - 2 weeks) solar PV generation forecasting. Provide a joint time series of solar and wind load factors together with synthetic transmission demand. Provide statistical characteristics of the relevant power (MW) and load factor variables, both on their own and jointly. Provide a concise library of extreme events for use in National Grid scenario modelling DELIVERABLE: A synthetic joint time series of solar and wind load factors at regional and aggregate GB level, together with transmission demand at National level and an accompanying report. This will provide details of the variability in solar PV load factors over a range of time and spatial scales, correlation with wind load factors and demand, and a library of interesting and extreme events. WP3. Assess the feasible improvements in short term (1-6 hours) solar generation forecasts. At short lead times, solar forecast errors could be reduced by making use of current observations other than Met Office solar irradiance observations. State of the art remote sensing data, satellite and radar, and new model products can be used to provide a better near real time spatial distribution of clouds and movement of frontal systems. Near-real time output estimates of solar generation can be blended with NWP forecasts. This work package will assess and develop statistical techniques for utilising this data feed. Both techniques can then be utilised in adapting the Met Office numerical weather prediction systems in between model runs. This work package will assess these data sources to improve short term solar PV generation forecasts in 1 - 6 hour time horizon. Use radar and satellite based remote sensing data options (e.g. radar) to improve NWP short term solar PV irradiance and generation forecasts Use near-real time data feeds of solar PV generation to develop statistical models to improve short term solar PV generation forecasts DELIVERABLE: A report summarising the improvements in short term solar generation forecasting that can be achieved using radar and satellite data sources, and from using near real time solar PV generation data feeds.Note : Project Documents may be available via the ENA Smarter Networks Portal using the Website link above
Publications	(none)
Final Report	(none)
Added to Database	11/12/18