Unibz Professors Francesco Ravazzolo (Economics and Management) and Maurizio Righetti (Science and Technology) cannot only help you make some sense of your electric bill, but they are also working on lowering the costs of producing energy, which could lead to lower prices for the consumer.
Take a glance at your monthly electric bill, and you will notice that your total consumption varies each month. Take a closer look and you will see that the average cost of each kilowatt-hour you consume changes too. A basic understanding of economic theory would tell us this is a simple matter of supply and demand. The electric power market, however, is particularly complex.
As the econometrician Francesco Ravazzolo and hydraulic engineer Maurizio Righetti explain, the electricity you use to power your home and workspace has been generated in several different power plants at several different locations, by means of several different methods, and using many different energy resources. Where and how your electricity was generated is constantly changing, which means the production costs of your electricity are constantly changing. At the same time, your consumption of electricity varies by time of day, day of the week, and season of the year. Add this all up, looking at aggregate supply and demand, and it is easier to understand why the price of electricity is so volatile.
The market is further complicated by the fact that demand for electricity is rapidly increasing (due to population growth, economic development, the accumulation of household appliances, etc.) but the demand for electricity produced with traditional fossil fuels (e.g., gas, oil, and coal) and nuclear energy is steadily decreasing. Environmentally conscious consumers are demanding more electricity produced with renewable energy sources (RES).
As Ravazzolo’s research demonstrates, the demand for and price of RES electricity is rather stable, but the supply is volatile due to dynamic weather and climate conditions. The supply of electricity produced with fossil fuels and nuclear energy is fixed in the short-run, but the demand and price fluctuate more than in the past due to the fluctuations in the supply of RES electricity. The shift in consumer behavior is therefore only making the market price even more difficult to forecast.
Like all businesses, electric power companies must plan future production in advance. But the complexity of the market for electricity makes it difficult to predict future market price and set production targets. And the stakes are rather high! When the demand and/or price is much lower than expected, the result is wasted electricity and a loss in profits for the producer. When the demand is much higher than expected, the result can be a shortage, which could result in power outages and additional costs, even fines.
In Italy, the Gestore dei Mercati Energetici sets the market price based on expected production and consumption. There are two common systems for calculating the market price. The first relies on a mathematical formula which forecasts future production capacities and consumption levels. The second is to rely on the prices indicated by a futures market. These methods however have a 10% margin of error, that is, the predicted value can be as much as 10% higher or 10% lower than the actual value. In order to prevent power shortages and outages, producers must maintain a “reserve margin” by intentionally overproducing.
Ravazzolo and Righetti are proving that we can better understand this complex market with big data and thereby optimize it, for instance by reducing the required reserve margin
The two professors—whose research is being funded by Alperia, the largest producer of electricity in South Tyrol—have teamed up to create an application which can more accurately predict the quantity of electricity supplied and demanded and the resulting market price one day in advance. To do this, they are collecting large amounts of data at a hydroelectric plant in South Tyrol to create a statistical model.
As Righetti explains, the reason they are developing an application for hydroelectric plants is because, unlike solar or wind energy, the energy potential of water can be stored in reservoirs and harnessed to produce electricity by simply opening the dam and activating a turbine with the flow. This means that hydroelectric plants can quickly respond to market fluctuations, that is, if operators know when and how long to open the dams.
Ravazzolo and Righetti are recording hourly observations across approximately 30 variables, including market price; consumption and production by source; imports and exports by source; weather conditions; water availability; and time cycles. Their model is constantly updated to include the previous four years of data, which means over 1 million parameters, or approximately 40 gigabytes.
The current margin of error of Ravazzolo and Righetti’s model is between 5% and 10%. The goal is to bring that down to around 5%.
The aim of their project is to optimize the level of production according to 1) expected market price given the dynamics of the market (Ravazzolo), and 2) the potential to produce hydroelectricity in the future given current technical capabilities of plants and the total availability of water in reservoirs and their catchment areas (Righetti). This optimization will enable energy providers to use hydroelectric plants more efficiently, in particular during peak hours of consumption, and thereby reduce the use of power plants than burn fossil fuels.
Righetti predicts their research will have a “waterfall” effect on all areas of life. Water is, in his view, our most “precious” resource. Hydropower accounts for the majority of electricity production in South Tyrol. Optimizing hydroelectric plants will also help us more efficiently manage and allocate water resources in household, industrial, and agricultural uses.
The second phase of their project will be to test and further develop their application at plants throughout the Euregio. The third phase will implement their application in all hydroelectric plants in northern Italy.