Using DLR to efficiently manage power lines
Dynamic Line Rating (DLR) is a technology used to dynamically adjust the capacity of electric overhead transmission lines. This capacity is determined by environmental cooling and is mostly dependent on weather elements such as air temperature, solar radiation, and wind speed and direction.
The latent transmission capacity made available by DLR means that more power can be transported through the power network. DLR can thus present a solution for enabling higher energy production whilst minimizing or postponing network reinforcements. The accommodation of extra power on the grid is highly beneficial. In the current power system landscape, random power injections from Intermittent Renewable Sources (IRS) can put stress on the existing infrastructure.
Until now, most Transmission and Distribution System Operators (TSO / DSO) manage their power network by applying static ratings. These ratings are based on conservative, seasonal assumptions of meteorological conditions around the power line. Unfortunately, DLR can be difficult to put into operation due to its variable nature compared to static seasonal line ratings. This could explain why DLR technology hasn’t taken off yet. In order to facilitate the integration, growth and practical application of DLR into the power system operations in The Netherlands, TenneT and MeteoGroup have jointly developed a weather-based thermal line rating model to produce more reliable DLR forecasts.
TenneT and MeteoGroup model
The model that MeteoGroup and TenneT have developed has evolved as a joint product of line engineering, operational knowledge, and meteorological expertise. Observations that were gathered during an 18 months pilot study have shown on average that a large capacity gain compared to the maximum capacity allowed by static rating. Cases of capacity over-estimation by the thermal line model were reviewed in terms of frequency, duration, and magnitude.
The output of the thermal line rating model can be expressed in probabilistic terms. This allows the network operator to set the desired forecast certainty level according to the operational requirements. Managing the Dutch power network through DLR requires reliable capacity forecasts one to two days in advance. The thermal line rating model is fed with state-of-the-art forecasts for wind speed and air temperature combined with their respective uncertainty bands.
Transport capacity and weather predictions
The transport capacity is calculated by solving the heat balance equation of the power line and expressed in relative terms as a percentage of nominal capacity. The nominal transport capacity is based on a maximum permissible line temperature of 80 C during conservative weather conditions: wind speed of 0.6 m/s perpendicular on the line, 35 C ambient air temperature and 900 W/m2 solar irradiance. The relative transport capacity is solely dependent on the weather conditions, irrespective of the technical characteristics of the conductor.
Those weather predictions which are used as input for the capacity model are based on spatially interpolating forecasts from a dense network of weather stations onto the individual masts of the power line. This ‘downscaling’ methodology makes use of high-resolution terrain roughness information to ensure that the wind speed is correctly interpolated onto the power line. The actual capacity prediction for the line is determined by the lowest capacity of all individual line spans.
The evolution and rating of the model
It is a challenge to rate the transport capacity of the electrical grid to a safe maximum. The TSOs / DSOs should meet the demands of the market as best as possible, without compromising on the stability and safety of the power network. For overhead power lines, this means that the maximum allowable sag is not exceeded, thereby violating the minimal distance of the line to the ground. The maximum allowable sag determines the maximum allowable conductor temperature and, with that, the thermal rating.
Real-time monitoring of the weather conditions at the power line or real-time monitoring of the measured sag are ways to obtain more accurate information in order to establish a safe thermal rating for that moment. Real-time solutions can help to manage the grid in an operational setting. However, since the network in the Netherlands operates under an N-1 principle, also during maintenance, there are very few moments that real-time rating would really help. In the Netherlands, the highest value of dynamic line rating can, therefore, be achieved with methods that accurately predict the rating some days in advance. This study produces such a rating in a reliable way.
A rating model has been developed to yield ratings for each hour of the day and at least two days ahead. This model is applicable all over the country.
Promising capabilities in the future
The concept of weather forecasted dynamic rating looks promising for application in the Netherlands. With hardly any hardware needed, it is possible to model all our 12500 towers and connect the predicted rating as needed to the programming and operating processes. In addition, intra-day programming can benefit from short-term forecasting dynamic rating. For sections in the network where an economic trade-off can be made between increasing year-round capacity and a probability of last minute or last hour measures to avoid overload, weather-based dynamic rating allows maximization of profits.
The thermal capacity of the overhead conductor is not always the limiting factor for a line. Network stability, voltage limitations and the rating of other equipment in substations, like circuit breakers, may also set limits.
For more information, download the report
At the Accenture Innovation Summit, November 2, in Utrecht, MeteoGroup will host a Climate RoundTable, featuring Tennet and titled: Optimizing grid capacity by providing better predictions for the early detection of congestion.
Register here for the Accenture Innovation Summit
Accenture Innovation Awards, Climate Round Table: Optimizing grid capacity by better predictions and early detection of congestion