AIS has worked in the field of forecasting demand in the food, editorial, financial and retail sectors. However, this issue is critical in the energy sector, where there are two aspects to consider: the forecast of consumption and incidents.
Domestic electricity consumption is mainly due to the use of large appliances whose use is seasonal: air conditioning, heating and appliances with an increasing presence in homes, eg dishwashers.To optimise the power supply required to estimate daytime and night-time consumption as accurately as possible.
The model developed by AIS can do the following:
- Explain current electricity consumption from different consumer profiles distributed over the territory: at the substation, by time slot and throughout the whole year.
- Predict future electricity consumption and simulate scenarios, both plausible and extreme, with the same degree of detail, based on the possession and use of household equipment and weather conditions.
- Provide a bottom-up approach for the electricity consumption model, from the individual consumption of different consumer segments in the territory, to explain current consumption.
Prediction of incidents in the power grid
A number of incidents occur in the process of distributing electricity through medium voltage lines and over time. Due to various factors, they often lead to power supply interruptions for a period for at least one client.
Some of the risk factors are controllable directly by the electric company (eg, the type of facility), while other factors are beyond their control (eg, the weather).
Examples of consequences of these incidents are loss of income due to interruption in the supply, loss of public confidence and complaints by customers adversely affected.
Faced with these problems, the aim is to improve the quality of customer service, reducing interruptions in the supply, both in frequency and in scope, in terms of energy not distributed, customers affected and duration.
AIS offers a solution consisting of the following phases:
- The first objective involves analysing and quantifying risk factors that have an impact on the frequency of incidents and their scope, distinguishing between those which are modifiable and those which are not. As a result of this study, actions to reduce the occurrence and extent of these incidents may be suggested.
- A predictive statistical model is built to determine the net contribution of each risk factor, and the expected and unexpected loss calculated from the probability distribution of incidents and their impact. These impacts from different investment campaigns can be evaluated at an individual level and different scenarios simulated.
- An optimisation model is constructed which includes the costs of various actions or investments to reduce the incidents. An optimum investment plan over several years can be prepared.
Subscribe to our newsletter
More information on our services and solutions