Electricity Load Prediction

LOADFOR (Load Forecasting) predicts the load in electrical power systems and delivers both highly competitive and accurate predictions. The maximal prediction horizon is determined by the meteorological forecast used and is typically one week.


LOADFOR is based on models developed for scenario analysis and long term forecasting, developed at the Technical University of Denmark over the period 1994-2000. Originally, the models were developed based on 20 years of hourly observations of power load and climate for a larger geographical region. Following this development the models were adapted for analytical purposes in cooperation with an electric utility. Finally the models were further developed and optimized for online operation, where the continuous flow of on-line observations are utilized.​

Input to LOADFOR​

LOADFOR takes the following data is taking into account:

Online measurements of power load.

Met. forecasts of the local air temperature, wind speed, and solar radiation for horizons up to the desired operational horizon, typically one week.

If available, climate measurements can be included also. Furthermore, by configuration, calendar information corresponding to the particular geographical region is used by LOADFOR.​

Output from LOADFOR​

LOADFOR produce online predictions of the power load for horizons up to the horizon covered by the meteorological forecasts, whereas the resolution in time can be less than that of the meteorological forecast.

As for other ENFOR forecast products the forecasts are supplied together with standard uncertainty intervals. For longer horizons where the meteorological forecast uncertainty may be significant the standard forecasts can be supplemented with quantile forecasts, which are highly reliable representations of the uncertainty.

In order to include the development over time in the representation of uncertainty reliable load scenarios can further supplement the standard forecasts. These scenarios can furthermore be generated jointly with other ENFOR scenarios as e.g. heat load or wind power production scenarios. In this way all relevant auto- and cross-correlations can be accounted for.​

LOADFOR as a web-service​

If desired, LOADFOR can be installed as a service on servers managed by us.​

Methods underlying LOADFOR​

LOADFOR is built on artificial intelligence, and hence the system automatically calibrates to the actual situation. For instance the models automatically take into account changes in the user profiles, the the consumer behavior around holiday periods, changes in the use of cooling, changes in the MET forecasts, etc.

The models are typically a mixture of parametric, semi-parametric and non-parametric models. As an example the effect of holiday periods are most often described by using splines, which allow for a high level of flexibility around the start and the end of a holiday period. Another example is the effect of the solar radiation (direct and diffuse) which most often is appropriately described by a non-parametric function, since the turn off of light as a function of the radiative components can not adequately be described by a parametric function. Some of the functions in LOADFOR are static functions, whereas as other functions are dynamic functions. As an example the influence of the outdoor air temperature is most often dynamically due to heat accumulation in buildings - provided that a part of the power is used for heating or cooling.


Below some further details regarding LOADFOR are presented.

Effect of holiday periods.

For predictions of the power load the effect of holiday periods (summer holidays, Easter, Christmas, etc) is most often one of the most complicated tasks. It has been found that the use of splines gives the most accurate predictions. Splines enables a high degree of flexibility by placing the knots around complicated time periods, but also by selecting the appropriate type of spline.


Day type effect.

Clearly there is a difference between work days and non-working days, but most often an effect for each day in the week is needed, since e.g. Monday is different from Tuesday, etc. Bank/public holidays, as e.g. New Year's Day, are assigned an appropriate day type which is not necessarily the same as the actual day of week.​​

Climate measurements / meteorological predictions.

The climate variables taken into account are: Ambient air temperature, wind speed, and solar radiation. The models are in general non-linear. Consider for instance the effect of outdoor air temperature (Ta) and the solar radiation R on the power consumption.

Figure 1 below shows contour plots of an non-parametric function estimate for the dependence of power load on outdoor air temperature and radiation. It is seen that a decrease in air temperature leads to an increased power consumption in particular for temperatures below, say 12 oC. For the solar radiation it is seen that below, say 80 W/m2, we see a dramatic increase in needed power for decreasing lower solar radiation, whereas for a solar radiation above 80 W/m2, the power load does not depend on the radiation.​​

Figure 1: Contour plot of the power load (in MWh/h) in Eastern Denmark as a function of ambient air temperature (Ta) and global radiation.

Thermal characteristics of the buildings etc.

The thermal mass of buildings implies that a rapid change of the outdoor air temperature does not lead to a similar rapid change in power consumption. This smoothing effect on the power load is seen with respect to both heating and cooling.

Consider the dynamic effect of air temperature and solar radiation. For a particular power system the estimated step response functions are shown in Figure 2. The step responses estimated shows that for the particular power system a persistent change of the air temperature is not fully effective until 3 to 4 days after the change occurred. For changes in the solar radiation the effect is seen to be fully effective after only 2-3 hours.​

Figure 2. Estimated step response functions on air temperature and solar radiation. The air temperature (Ta) dropped from 0 deg.C to -1 deg.C at t=0, and the global radiation (Rg) dropped from 25 W/m2 to 20 W/m2 at t=0.​

​Contact information


Lyngsø Allé 3

DK-2970 Hørsholm

​​Phone: (+45) 45 350 350

E-mail: info@enfor.dk

VAT no.: 29421633

Lyngsø Allé 3, 2970 Hørsholm
Tlf.: 45 350 350