By 2035, it is expected that more than 700 GW of production capacity will have to be built in Europe, mainly due to the retirement of existing power plants, the widespread use of renewables and the construction of additional facilities for fossil fuels to ensure the reliability of the of the power system operations. Also elsewhere around the world, production capacity will be built between 1960 and 1970, during the period of high growth of electricity consumption, new capabilities will be introduced, leading to more than 5,000 GW of investments by 2035, of which more than half of renewable energy (wind , sun, water, biomass).
As electricity consumption is constantly increasing, although slower in recent decades, there is virtually no uncertainty about the need to build new capacity. One of the biggest uncertainties in this area is whether political measures will lead to a situation where most of the new production facilities will be subsidized or not. is regulated in some other way (mainly from renewable sources). In addition, the most common renewable sources, such as the wind and the sun, are very unstable and therefore we need classical power plants that ensure the reliability of the system. In general, the uncertainty about the need for new conventional energy sources is rather low and apparently it will be mostly thermal power plants, as hydro potential is already largely exploited.
Nevertheless, there is considerable uncertainty about which power stations will be most cost-effective during the period considered. This is partly due to the uncertainty of the prices of primary energy products. In particular, with the increase in gas prices since the end of the twentieth century, it became rather questionable whether the natural gas plants are adequate. The power plants on coal, lignite and uranium are the following options. However, when selecting power plants for construction, we must take into account a number of uncertainties. The optimum choice of power plants is influenced by political, technical, economic, social, environmental and other uncertainties such as, for example, changes in environmental legislation, decisions to close nuclear power plants, future growth of electricity consumption, technology development.
These uncertainties require new methods to assess the eligibility of the construction of various types of power plants. A simple extrapolation of current patterns of consumption, prices and other variables in the future seems problematic because the current situation due to significant overcapacity in Europe can barely reflect the potential shortage of production capacity at the closure of older power plants. Therefore, the future state of the EES (consumption, price, spare capacity, etc.) must be considered simultaneously with the investments and retirements of existing power plants.
Particular attention must be paid to electricity generation companies, because in addition to being exposed to a number of uncertainties brought about by the free market in electricity, investments in electricity generating facilities are very specific.
Using a mathematical model, we tested the applicability of multiobjective optimization techniques with evolutionary algorithms under uncertainty in case of a decision to invest in a new electricity generation facility. We analyzed the properties of multiobjective decision-making in energy using evolutionary algorithms, the Bayes approach to estimating the parameters of statistical models, and the function of the cube for modeling the probability laws of dependent random variables. Subsequently, we combined this into a model of a stochastic multiobjective optimization of the development of generation sources under uncertainty, which we tested under various scenarios of the development of the electric power system.