Hourly electricity price forecasting : A case study of Italy 

Get Complete Project Material File(s) Now! »

Deregulation of electricity market in Italy

From 1960’s, Italian electricity market was controlled, managed, and organized under a solitary vertically integrated company, Enel, which is owned by the state. Therefore, generations, transmissions, and distributions are all operated by Enel. However, with new EU directive on energy market deregulation was passed in 1996, many changes were applied in Italy. In 1999, legislative decree 79/99 was agreed by the parliament and has sparked the start of the free electricity market in Italy. 1. In line with the European directives, the decree, now widely known as Bersani decree, address many issues regarding liberalization, renewable dispatch, and green certi cate.
The liberalization process starts with ending the national monopolies and opening the market for new entrants in order to promote competition. The decree provided speci c requirements to Enel for re-structurization of the electricity sectors. The requirements are:
Enel had to start privatization and to unbundle all their activities into di erent accounts. The activities are :
{ Power generations
{ Energy distributions and sales
{ The dismantling of nuclear power plants
Transmission activity in the electricity sectors must be given up to an indepen-dent body with non-discriminatory rules.
Enel had to sell at least 15 GW of their energy capacity in order to reduce their market dominance and to open electricity market for new entrants.
The transmissions activities, as dictated by EU directive, was designated to a new state-owned company, Terna, which was a part of ENEL group. Later in 2003, they had to be separated from Enel as a consequence of new EU directive number 54. Now, they are an independent company owned by the Italian government and private investors. As for the sales of capacities, the government facilitated Enel for the best method of these sales in order to guarantee fair market conditions and transparency. The generation units were, then, sold to major EU companies. The new owners of these capacities are:
Endesa, Spanish utility company, and the bank of Santander, major Spanish bank, who jointly acquired 5.34 GW
Edison, Italian utility company, and Atel, swiss utility company, who jointly bought 7 GW of the capacity.
Energia Italia, Italian utility company, and Electrabel, Belgium utility company, who obtain 2.6 GW from the sales.
As the decree suggested, the next phase of the deregulated market was demand liberalization and creation of internal energy market, Italian government started to open the market and considered a further measure to ensure the implementation. The market was initially opened for large industrial customers that consume 30 GWh of electricity every year. Therefore, there was still large portions of customers that did not have the option to choose their energy suppliers. This is, later, improved gradually between 1999 until 2007 as follows :
in 1999, users with a minimum 30 GWh/year of consumption is considered as eligible customers
in 2000, users with a minimum 20 GWh/year of consumption is considered as eligible customers.
in 2002, users with a minimum 9 GWh/year of consumption is considered as eligible customers.
in 2003, users with a minimum 100 MWh/year of consumption is considered as eligible customers.
in 2004, all non-household users is considered as eligible customers in 2007, all users are considered as customers.
As far as the creation of internal energy market concerned, operators were ini-tially able to trade electricity through bilateral contracts. However, the decree also recommended the establishment of the independent market operator that operates the market with the objective to fully implement EU directive. The establishment of an independent body to manage the market, operators were expected to be able to buy or sell electricity by two means, bid/o er of standardizing contracts in the electricity exchange and over-the-counter contracts. Therefore, this decree initiates the legal framework for the creation of Italian power exchange and the formation of the Gestore Mercato de Energi (independent market operator) in 2004
It is also interesting to be noted that the Bersani decree also initiates green cer-ti cate mechanism which was designed as a change for feed-in price policy under CIP6/92. In this mechanism, instead of xed guaranteed income, the new support is traded base in the market under the quota system. The quota system obliges power producers and importers to produce a certain quota of their outputs from renewable sources. The quota starts from 2% and gradually increasing annually. Green cer-ti cates were used to ful ll this obligation as they are able to purchase them from third parties. The certi cates were traded on a parallel market, independent of the electricity market. Then, as encouragement for renewable energy penetration, the de-cree provided dispatching priority to its production and increase government support for their researchers.2In addition,it also introduced a renewable-energy quota system. Several years later, improvements had been made to the green certi cate scheme in order to promote renewable-energy further. In 2007, under the Law 244/2007, small generators (0.2 MW -1 MW capacity) have the option to obtain a support scheme by selling their green certi cates on the market or receive a feed-in tari . In addition, the law also extended the period for the release of green certi cates up to 15 years for new and refurbished installations. In order to enhance renewable growth, nance act of 2008 and 2009 ministerial decree increased the quota by 0.75 % per annum over the years between 2007 and 2012. Consequently, power producers and importers were required to produce 6.8% of their supply from a renewable resource in 2011 and 7.55 in the later year. The law also introduced another mechanism that considers the technology maturity. Hence, the quantity of the certi cate granted to producers and importers, with more than 1 MW capacity, was multiplied by a coe cient depending on the technology ranging from 1.0 for the onshore wind turbine to 1.8 for wave en-ergy conversion. Moreover, according to the law, one certi cate had a contract size of one MWh instead of 50 MWh as in the previous regulation. Therefore, it can support newer installations for small renewable-energy technology. In 2011, a new legislative decree was passed on 3 March. The decree set a national target of 17% of renewable energy in the gross nal energy consumption by 2020 in compliance with European Directive (2009/28/EC). With a view to realizing the goal, the decree changes the quota system back into a feed-in tari scheme from 1 January 2013 under a given threshold and a tendering scheme for new plants (except biomass) with a capacity above the threshold. The threshold di ers according to the type of technology. There-fore, a guaranteed xed price is o ered to support the renewable energy penetration. Consequently, GSE was obliged to buy all certi cates that exceed the annual demand in 2012.


The deregulation of electricity markets in Italy has changed the shape of the economy in the electricity sector. Furthermore, speci c characteristics of Italian electricity market makes this particular electricity market an interesting case study. Firstly, Italian governments and independent body in the electricity market have decided to apply an inter-zonal pricing mechanism for the price formation. The proposed price mechanism would not only re ect actual supply and demand in Italy but also re ect the actual net physical exchange between their administrative zones. This pricing mechanism was a normal practice as it has been applied in many countries such as United States, Australia, and Denmark. However, in comparison to other countries, Italy has the largest number of zones in their national energy market reaching six zonal prices. Secondly, recent ambitious policy by Italian government has increased the intermittent generations in the exchanges. The dependency on the weather has created uncertainty in the supply. As a result, the price has become more volatile compared to previous years. This is a big problem from the point of vies of the operators as they are exposed to nancial risk. Thirdly, capacity transmission lines between Italian zones are not equally distributed. As the renewable generation units are located strategically to harvest energy, they are generally far from the population where there is a high demand for electricity. As a consequence, congestion occurrence becomes a problem in Italy since the infrastructure is not equal in all the zones.
These interesting facts in Italy have brought our attention towards analysis on the Italian electricity market. They in uence our motivation to explore the literature with the aim to seek the gaps in academic and to contribute to the economic and empirical research. Our studies have brought us to three main line of researches that motivates us to explore intensively in this particular market. They are :
Alternative models to forecast Italian electricity market
The impact of renewable supply on congestion and congestion cost
Examination of national integration in the Italian electricity markets.
The following subsections are dedicated to brie y describe the literature gap in these line of research and its motivational background.
Alternative models to forecast Italian electricity market
The electricity price has changed its shape after the electricity market liberalization and has sparked many researches in this eld. This is mainly due to the fact that previous electricity price was not volatile since it was xed through policy. In fact, all the activities in the Italian electricity sector were controlled, managed and organized by one state-owned company. This old framework provides no competition in the market and creates an ine cient economy. The deregulation, on the other hand, has opened new entrants in the electricity sector and has provided a fair price to re ect the actual supply and demand. The new mechanism has given the opportunity to new producers and energy retailers to compete for the right to deliver/come electricity in the exchange. Furthermore, the price would re ect the actual scarcity and surplus of electricity in the market based on the actual supply and demand.
Despite the bene ts o ered by the market, there were issues that come because of the uncertainties provided in the market. As the price is formed by supply and demand functions, deviations on these functions would create shocks in the market equilibrium. These shocks can be negative because oversupply in the market, which would translate into a low price, and can be positive because a scarcity in the market, which would translate into a higher price. As a consequence, the price can changes rapidly from zero (or negative) to a price jump reaching more than 100 AC/Mwh. This is also due to the fact that electricity cannot be stored thus becoming vulnerable to price spikes. In Italy, the rise of intermittent energy in the power exchange has worsened the situation as it increases the volatility in the electricity price (see for instance Clo et al., 2015 ). On the other hand, Italy has the highest average wholesale price in comparison to the other mature EU market since gas is still their marginal technology (GME). Hence, it indicates a high deviation in the electricity price caused by demand changes, fuel price changes and supply deviation. These price changes exposed the market participant to nancial risk in the electricity market. The changes in the spot price can reduce their economic gain because of the lost from price fall or jump. On the other hand, policy makers are obligated to follow the fair price in the market. Therefore, it is necessary to provide a reliable model to forecast electricity market that could also re ect the power exchanges.
Our research focused on research on the best forecasting model under statistical class 3 with the main objective to contribute to the limited literature of Italy elec-tricity market. Our research o ers an alternative model for forecasting hourly price in Italy since there is only a few researches in this particular eld and market. The statistic class was chosen because it has the advantage of having an accurate quan-titative prediction, having less required observations and having results that can be interpreted in an economic sense. Based on our review, we can only nd four alter-native models, under the statistical class, to forecast electricity price in the Italian electricity market’s literature. 4 The four models existed in the literature are Pe-riodic Autoregressive models (Bosco et al , 2007) ARMAX-GARCH (Petrella and Sapio, 2011), Generalized Additive model for Location, Scale, and Shape (Serinaldi ,2011) and ARFIMAX-GARCH, (Gianfreda and Grossi, 2012).

READ  Board independence and the monitoring-advising trade-off 

The impact of renewable supply on congestion and congestion cost

As it was mentioned in the previous subsection, EU directive number 28 in 2009 (20-20-20 policy) has set ambitious goals to show commitment to combating climate change. As a consequence, it has ignited the interest in renewable energy investment in Europe because many EU member state starts supporting schemes in order to meet their national renewable energy target by 2020. In 2013, the renewable energy generations in Europe has supplied 14.95% of gross nal energy consumption. These changes have motivated researchers to study the impact of the renewable energy on the electricity markets. Economic literature has put their main focus on the changes in wholesale price. These researches have underlined two main changes in the wholesale price, the reduction in terms of price level and the increase in volatility. The empirical evidence of these two phenomena has been well-documented in the academic literature with various case study and methods. 5
However, the rise of renewable supply has raised another issue in regards to the electricity network, which has not intensively explored in the academic literature. As renewable production units are generally located far from the demand site because of optimization in harvesting energy, high capacity transmissions are required to ensure electricity delivery. However, the existing network was not built for high renewable energy penetration, thus, the changes in the renewable energy mix, consequently, put additional stress on the infrastructure, amplifying transportation needs and multi-plying congestion occurrence. Unfortunately, there are not many works of literature focused on this particular issue.6
Our studies are aimed to contribute to the limited literature in this particular eld. Based on our knowledge, Sapio (2015) is the only author who studies the impact of renewable to the congestions in Italy. Utilizing regime-switching model, he estimated Italian power exchange data between 2012 and 2013 with a focus on Sicily and Southern Italy connection. The paper concludes that rising renewable production in Sicily reduce the congestion as the zone require less import from its neighbor. The increase of renewable from outside Italy (Southern Italy), on the other hand, would increase the urgency to export the electricity to its neighbor. Consequently, it results in the increase of congestion towards Sicily. The paper also concludes that wind energy supply provide a bigger impact on congestion in comparison to solar productions unit.

Isolated regional market

Deregulation of the Italian electricity market has changed the previous mechanism into a cost e ective mechanism for electricity generation and transmissions. This was one of the top priority from the policy maker since an e cient market and network was the main objective of the EU directive and Bersani decree. In order to reach cost ef-fective transmission cost, Italy adopts inter-zonal pricing mechanism which separates national electricity market into several zones based on the conditions of the electricity system. If a congestion was found in the system, the mechanism allows zonal market prices to be higher in order to balance the system and pay the cost of physical delivery between the zones. However, this could be a problem for countries with low capacities on inter-zonal transits since it will increase the congestion frequency. Consequently, the national markets would constantly splits, which results in isolation of zonal mar-kets. Therefore, regions that do not posses adequate network capacity for physical exchange with its neighbor would fail to integrate into the national market.
This issue motivates us to examine the integration of Italian electricity market that is never explored by previous literature. Having six zonal markets on their national market, Italy o ers interesting study for our research. Furthermore, new installations of transmission lines between Sardinia and Italian peninsula provide us with a unique case study for examining its impact on the regional market integration. We may examine the integration by studying the interdependencies of zonal prices since it re ects the linear relationship among the regional markets. Strong interde-pendencies suggest a stable transmission line and a full integration in the national market. Previous literature on this line of research is mainly used in a case study from outside Europe, Australia, and United states. In Australia, Worthingthon et al (2005), Higgs (2009) and Ignatieva and Truck (2011) have explored interdepen-dencies of ve regional markets using di erent methods. The researches provide a general conclusion that regional markets with better networking infrastructure dis-play strong interdependency whereas the weaker level of interdependencies is recorded in the markets with low capacity of the transmission line. In the United States, Park et al (2006) analyze United states spot markets using Vector autoregression. Their papers conclude that interzonal transits and national market organization a ect the interdependencies. Another national market examination in the United States was done by Dempster et al (2008) for California electricity market by means of granger causality test.

The Italian electricity market

Five years after the Bersani decree passed, the national electricity exchanges could be nally operated. The rst stage of IPEX commenced on 1 January 2004 as technical trials with operators from supply and demand sides. The technical trials were aimed to ensure a proper infrastructure and mechanism have been installed in order to ensure a transparent and competitive electricity market in Italy as well as to avoid the risk of an unbalanced system. The seconde stage of IPEX was partial functioning of the exchange (only for a supply side) as the initial trials. Then, the beginning of 2005 was a milestone in Italian electricity sector as Italian power exchange starts complete operation of the exchanges. The market is now widely known as Italian Power Exchange (IPEX). Currently, it has 254 market participants who actively buy and sell electricity contracts with a total of 337.34 TWh of electricity was traded in 2014.
Beyond 2009, Italian power exchange did not go through any major structural changes. The market had matured enough to operate regularly in the transparent fashion, without the needs to majorly modify the structure for improving liquidity and risk mitigations. Between 2005 and 2009, Italian power exchange has seen three major structural changes that a ect the exchange. The major structural changes are :
Demand liberalization
As it was mentioned in the previous literature, Italy adopted gradual demand liberalization between 1999 and 2007 that a ects the structural changes in the market. When IPEX was opened and fully operated in 2005, only industries were allowed to enter a bid in the exchanges. Hence, all households consumers were not considered eligible by the legal framework. However, this is changed after July 2007 as all consumers become eligible in the power exchanges. This change has put 85 TWh of yearly demand in the markets which constitute 22 million families. Consequently, this change would introduce high market uc-tuations in the short-term electricity price in Italy thus increasing volatility. Furthermore, these new entrants may engage in a trial-and-error learning pro-cess that produces instability in the market.
Contract for Di erences
Starting from 2008, the ensuing price risk from the day-ahead market in IPEX can be hedged by contracts for di erences (CfDs), which is held in Italian derivatives exchange (IDEX) operated by Borsa Italiana. As a result, hedg-ing strategy can be implemented by market participants. In fact, Acquirente Unico is obliged by law to hedge against price and volumetric risks. This ma-jor structural changes reduced market participants’ exposure towards nancial risk. However, before December 2009, holders of the contracts are not able to physically exercise the nancial contracts that they obtain from IDEX. Hence, GME, nally, introduces Piattaforma Consegna Derivati energia (CDE) with the purpose of converting the nancial derivative contract concluded in IDEX to physical delivery of electricity.
The last major structural changes in the Italian structure was done by the end of 2009. Under art 17 of Annex A to AEEG’s Decision 111/06, a special platform called Piattaforma dei Conti Energia a Termine (PCE) is started to be operated and managed by GME in order to facilitate electricity delivery from bilateral contracts concluded outside the IPEX. The platform is used by participants to register their injection and withdrawal schedules thus easing the works for managing transmission. This change has provided the market participant with options to choose between standardized contracts in physical exchange or OTC contracts. Furthermore, the platform o ers the posibility to adjust easilly both long term contracts and short term contracts. As a result, the markets becomes more active and competitive which a ects the Italian electricity prices.

The exchanges

The Italian electricity market consists of two classi cations, spot, and forward market. The market structure can be seen in the Figure 1.1 below. All markets are organized by the Italian independent market organizer (GME) 7 with an exception for nancial derivative contracts, which is traded by Italian derivatives exchange (IDEX) of Borsa Italiana. The IPEX organizes all the markets under GME SpA with an exception for nancial contracts, which is traded in IDEX by Borsa Italiana. GME also manages the OTC registration platform, PCE, where all market participants can register their bilateral contracts concluded outside the IPEX or IDEX. Our analysis and studies, however, focus on solely on the day-ahead market where 24 periods of electricity delivery are traded. In Italy, the day-ahead market opens on the ninth day before the day of delivery and close on the day before the delivery. Between this period, all buyers and sellers are allowed to make bids for their willingness to consume/produce electricity by specifying the quantity and the maximum/minimum price they desire.
8 After the market closes, an authomatic algorithm is used optimize the transmission system and determine the electricity price.
In order to maintain and optimize Italian electricity system, IPEX divides Italy into six main geographical zones and ve poles of limited production. The map of Italy in Figure 1.3 below displays zonal division of all the twenty administrative regions in Italy. NORD has the biggest territory covering eight regions followed by SUD with four. Sicily and Sardinia are the only two regions that are considered as one zonal market. The inter-zonal connections of the Italian electricity system can be found in the Figure 1.3 below. The electricity markets are coupled with four foreign markets, France, Switzerland, Austria, and Slovenia. Hence, it allows to import/export to/from Italy. The six main zonal markets are connected to the pole of limited productions or constrained zones. These zones are only a sets of generation units connected to Terna without withdrawal points (no demand in this limited zone). The maximum production that can be exported to the grid is lower than the maximum net physical exchange in the case of congestion. In other words, they are used to to balance the Italian electricity system in order to avoid congestion.

Table of contents :

1 Introduction 
1.1 Background and motivation
1.1.1 Deregulation of electricity market in Europe
1.1.2 Deregulation of electricity market in Italy
1.1.3 Motivation
1.2 The Italian electricity market
1.2.1 The exchanges
1.2.2 Liquidity
1.2.3 Price formation mechanism
1.2.4 Production Mix
1.2.5 Physical exchange
1.2.6 Transmission limit
1.2.7 Zonal price dierence
1.2.8 Congestion
1.2.9 The database
1.2.10 Stationarity
1.2.11 Electricity Prices
1.2.12 Demand
1.2.13 Renewable energy supply
1.3 Three angles of Italian electricity market
1.3.1 Hourly electricity forecast
1.3.2 Impact of renewable supply on congestion
1.3.3 Interdependency of Italian electricity market
2 Hourly electricity price forecasting : A case study of Italy 
2.1 Introduction
2.2 Empirical framework and analysis
2.2.1 Univariate framework
2.2.2 Evaluation Method
2.2.3 Univariate Model evaluation
2.2.4 Main fundamental drive
2.2.5 Multivariate Framework
2.2.6 Empirical Result and analysis
2.3 Conclusion
3 Intermittent renewable generation and network congestion: an em- pirical analysis of Italian Power Market 
3.1 Introduction and literature review
3.2 Empirical strategy
3.2.1 Multinomial logit model
3.2.2 2SLS with segmented regression
3.3 Conclusion
4 Interdependency of Italian zonal prices 
4.1 Introduction
4.2 Model Specication
4.2.1 First Regression
4.2.2 Second Regression
4.3 Results
4.4 Conclusion
5 Conclusion 
5.1 Main conclusions
5.2 Future research direction


Related Posts