Article Impact on Electricity Consumption and Market Pricing of Energy and Ancillary Services during Pandemic of COVID-19 in Italy
Emilio Ghiani * , Marco Galici, Mario Mureddu and Fabrizio Pilo
Department of Electrical and Electronic Engineering, University of Cagliari, 09123 Cagliari, Italy; [email protected] (M.G.); [email protected] (M.M.); [email protected] (F.P.) * Correspondence: [email protected]
Received: 24 April 2020; Accepted: 19 June 2020; Published: 1 July 2020
Abstract: At the moment of writing, in Italy, there is an ongoing pandemic of coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Its outbreak is leading to severe global socioeconomic disruptions impacting on all economic sectors from tourism, industry and the tertiary sector, up to the operational and opening of public offices, the closure of schools and the organization of families. Measures adopted by the Italian government to deal with the COVID-19 emergency have had direct effects both on people’s daily lives and on the activity of most industrial and commercial production companies. These changes have been unequivocally reflected also on the Italian electricity system, which has shown unprecedented behavior in terms of both energy consumption and volume—and subsequently, in the observed share of renewable and conventional production technologies. The goal of this study is to show the impact on the power industry of all the restrictions and lockdown of the activities in Italy and to discuss the effects of COVID-19 outbreak on the bulk power system and the entire electricity sector. In particular, the consequences on load profiles, electricity consumption and market prices in Italy, including the environmental aspects, are examined.
Keywords: pandemic COVID-19; bulk power system; electricity consumption; load profiles; day-ahead market
1. Introduction The coronavirus SARS-CoV-2 emerged in Wuhan, a city of 11 million people in China’s Hubei province in late 2019. Cases of the COVID-19 disease grew by several thousand per day in China in late January and February 2020 [1] since then, the outbreak evolved in a pandemic centered in Italy, with major outbreaks in South Korea, Iran and the US. In particular, the northern Italian regions of Lombardia, Veneto and Emilia-Romagna have been most affected by the outbreak, with thousands of infected patients and deaths. In an attempt to slow down the infection, the Italian government started a nationwide lockdown on 8th March 2020, with huge economic repercussions on the country. During the lockdown, the government imposed countrywide precautionary measures. This involved not only the forced quarantine for the population, closing of schools and churches, shops, bar, restaurants, retail shops, but also shut down of national and international flights, closure of non-essential ports and airports and non-essential firms and industries. This caused a reduction in energy consumption, leading to a decrease in the overall demand in the energy market. The COVID-19 measures in Italy quickly started to cause a noticeable impact on energy consumption, which reflected in daily profiles and market prices. Notably, the forced closure of industries and tertiary activities highly flattened the working hour consumption curves in comparison
Energies 2020, 13, 3357; doi:10.3390/en13133357 www.mdpi.com/journal/energies Energies 2020, 13, 3357 2 of 19 with the daily pre-lockdown ones. Although many people started home working through smart working, the increasing of residential consumption did not compensate for the general reduction of demand. As a result, the daily consumption curve is now less pronounced than a typical spring day. In the morning, electricity consumption ramps up slower, since there is no morning commute. The evening peak is still observable, as people prepare meals, turn lights on and spend time watching TV and playing games—but it is much lower than would normally be expected. The COVID-19 outbreak on a global scale is deeply influencing the whole energy industry, causing an overall reduction of oil and natural gas demand which, coupled with strategic tensions among the leading producing nations, led to an unprecedented reduction of their selling prices. BRENT oil price plummeted about 59% from the last peak reached on 17 of February (40 days before the writing of this study) to a minimum value of USD 24 per barrel, an all-time minimum since 2002. Similarly, natural gas prices fell to USD 1.63 per million metric British thermal units (mmBtu), showing a relative decrease of 38% from the November 2019 peak, reaching its minimum since November 1995. The combination of both local and global effects caused a strong reduction in electricity demand prices, which is going to influence the entire electricity supply chain in the short and middle term [2]. In addition, the IEA reports forecasted a global yearly total demand decrease of 6%, especially hitting the production from non-renewable energy sources [3]. Europe, moreover, showed a consumption decrease of roughly 10% during the final weeks of March and the first week of April [4]. This decrease has been observed to be particularly severe in Italy, which reportedly reached a decrease in consumption of about 23% [4] as described also in the IAEE Energy Forum [5] which analyses worldwide the impacts of COVID-19 on the energy industries in term of demand drops, supply-side shocks, facilities shutdowns and new patterns of electricity demand. Due to the aforementioned reasons, the effects associated with the COVID-2019 outbreak are difficult to assess with the traditional evaluation of the performance of critical infrastructures, based on classical risk assessment methods, which consider credible and predictable hazards and threats [6]. The ubiquitous effects on the power system should be indeed studied from a resilience perspective. Resilience definitions and frameworks have been proposed in different disciplines and applied in a variety of case studies worldwide [7]. In most resiliency studies, energy systems are subject to major disruptions affecting economic activities, operation of infrastructure and the society as a whole [8]. Even though, in this case, the resiliency issues arise from a crisis that has occurred in civil society and impacted the operation of the Italian power system and its related energy market prices. Thus, the data relative to the COVID-19 outbreak period will be of vital importance for testing the reliability of the currently employed planning, scenario-making and resilience testing methodologies. In particular, accurate analysis of the current situation will allow a better estimation of scenarios with higher penetration of renewable energy sources (RES) and their effects on the future power industry. In this context, the goal of this study is to provide detailed and comprehensive information related to the alterations on load profiles, electricity demand and both wholesale and ancillary services market prices in Italy due to the restrictions and lockdown of the activities, providing powerful insights about the resilience of the entire electricity sector for a prolonged period of low consumption/production ratio and posing issues which can serve as inspiration for future research.
2. Generation and Consumption in Italy before the COVID-19 Outbreak
2.1. Generation The Italian electrical system is the energy backbone of the third-largest economy of the European Union. In 2018, the Italian electricity system had a peak consumption of around 60 GW and a generation capacity of 115.20 GW [9]. As depicted in Figure1, the generation capacity in Italy consists mainly of thermal generation (coal, gas, oil) and renewable energy sources (RES); nuclear production is forbidden by law since 1987 and about 15% of the electricity is imported from abroad. Energies 2020, 13, 3357 3 of 19
RES consolidated their leading role in the Italian energy system, confirming themselves as a Energiesdecisive 2020, 1 factor3, x FOR for PEER the REVIEW sustainable development of the country in the energy transition. According3 of 22 to Italian national energy strategy (NES) [10], the integrated national energy and climate plan [11] and Italianthe nationalnational electricityenergy strategy transmission (NES) [10], grid the development integrated plannational [12], energy RES capacity and climate is expected plan [11] to grow and up theto national 93 GW electricity in 2030, with transmission almost 40 grid additional development GW with plan respect [12], RES to 2017. capacity The is main expected contribution to grow will up be to 93from GW solar in 2030, photovoltaic with almost (PV), 40 with additional almost GW+30 with GW, followedrespect to by 2017. wind, The with main almost contribution+9 GW. will be from solar photovoltaic (PV), with almost +30 GW, followed by wind, with almost +9 GW. 140
120
100
80
60
Capacity [GW] Capacity 40
20
0 Thermal Hydro Photovoltaic Wind Geothermal Installed capacity
FigureFigure 1. Italian 1. Italian power power capacity. capacity. The actual gross electricity production from RES in summarized in Table1[ 13]. Thermal generation The actual gross electricity production from RES in summarized in Table 1 [13]. Thermal is fueled by natural gas sources (72.6%); whereas coal and oil account for 20.9% and 6.5%, respectively. generation is fueled by natural gas sources (72.6%); whereas coal and oil account for 20.9% and 6.5%, respectively. Table 1. Gross electricity production (TWh) from renewable energy sources (RES) in Italy compared to total load. Table 1. Gross electricity production (TWh) from renewable energy sources (RES) in Italy compared to total load. Source 2015 2016 2017 2018 2019 SourceHydro 2015 46,436 2016 43,771 2017 37,545 2018 49,918 2019 47,063 Wind 14,705 17,522 17,565 17,556 19,992 Hydro 46,436 43,771 37,545 49,918 47,063 Solar 22,438 21,629 23,864 22,121 24,138 GeothermalWind 14,705 5814 17,522 5865 17,565 5784 17,556 5707 19,992 5688 SolarBiomasses 22,438 17,180 21,629 17,212 23,864 17,010 22,121 16,896 24,138 16,881 GeothermalTot RES 5814 106,573 5865 105,999 5784 101,768 5707 112,198 5688 113,762 BiomassesThermal 17,180 162,55 17,212 168,906 17,010 177,497 16,896 160,822 16,881 163,558 Tot LoadRES 106,573 316,95 105,999 314,3 101,768 320,578 112,198 321,500 113,762 319,552 Thermal% RES 162,55 33.6 168,906 33.7 177,497 31.7 160,822 34.9 163,558 35.6 Load 316,95 314,3 320,578 321,500 319,552 2.2. Load Profile of the% ItalianRES Power33 System.6 33.7 31.7 34.9 35.6
2.2. LoadFigure Profile2 ofshows the Italian the di Powerfferent System contributions of consumption for each sector in Italy during the year 2018. The main consumption comes from the residential load (39%). The secondary (industrial) and tertiaryFigure (services)2 shows the sectors different have 37%contributions and 22% of totalconsumption load, respectively, for each sector whereas in Italy the primary during the sector year(agriculture, 2018. The main raw materials)consumption account comes for from 2% ofthe total residential demand load [9]. (39%). The secondary (industrial) and tertiaryThe load(services) profile sectors is the shapehave 37% of the and electrical 22% of loadtotal overload, time. respect Theively, load whereas profile forthe a primary given area sectorvaries (agriculture, considerably raw according materials) to account the categories for 2% of of users, total demand including [9]. commercial, residential or industrial users. Peak electricity demands also vary according to seasonal and holiday needs. A typical Italian
Energies 2020, 13, x FOR PEER REVIEW 4 of 22
Energies 2020, 13, 3357 4 of 19 140
120 load profile in 2019, for winter (January) and summer (July) weekdays and weekends, is shown in Figure3. Electricity100 consumption normally follows a regular daily shape; quick rising as people wake up and go to school or work, smoothening throughout the day and then peaking at dinner time when people get back home.80 In Italy, the flattening of the load curve during the daytime, as shown in Figure3b, is exacerbated in summer by the production from PV distributed generation in distribution 60 Energiesnetworks 2020, [ 1413,, x15 FOR]. PEER REVIEW 4 of 22 40 140
Energy Consumption [TWh] 20 120 0 100 Primary Sector Secondary Tertiary Sector Residential Sector Sector 80 Figure 2. Subdivision among sectors of the energy consumed in Italy. 60 The load profile is the shape of the electrical load over time. The load profile for a given area varies considerably40 according to the categories of users, including commercial, residential or industrial users. Peak electricity demands also vary according to seasonal and holiday needs. A
typical ItalianEnergy Consumption [TWh] load20 profile in 2019, for winter (January) and summer (July) weekdays and weekends, is shown in Figure 3. Electricity consumption normally follows a regular daily shape; quick rising as people wake up 0and go to school or work, smoothening throughout the day and then peaking at dinner time when peoplePrimary get back Sector home. SecondaryIn Italy, the flatteningTertiary of Sectorthe load curveResidential during the daytime, as shown in Figure 3b, is exacerbated in summerSector by the production from PV distributedSector generation in distribution networks [14,15]. FigureFigure 2. 2. SubdivisionSubdivision among among sectors sectors of of the the energy energy consumed consumed in in Italy Italy..
60 60 The load profile is the shape of the electrical load over time. The load profile for a given area 50 varies50 considerably according to the categories of users, including commercial, residential or industrial40 users. Peak electricity demands also vary40 according to seasonal and holiday needs. A [GW] typical[GW] 30 Italian load profile in 2019, for winter (January)30 and summer (July) weekdays and weekends, Load Load is shown20 in Figure 3. Electricity consumption normally20 follows a regular daily shape; quick rising as people wake up and go to school or work, smoothening throughout the day and then peaking at 10 10 dinner time wheWorkingn people Day Winter get backHolyday home. Winter In Italy, the flattening Workingof the Day load Summer curve duringHolyday Summerthe daytime, 0 0 as shown1 2 in3 Figure4 5 6 7 3b,8 9 10is11 exacerbated12 13 14 15 16 17 18 in19 summer20 21 22 23 24 by the1 production2 3 4 5 6 7 8from9 10 11PV12 13distributed14 15 16 17 18 19 generation20 21 22 23 24 in distribution networks [14Time,15 [h] ]. Time [h] a) b) 60 60 FigureFigure 3.3. LoadLoad profileprofile ( a)) Weekday/Weekend Weekday/Weekend (January/Winter) (January/Winter) (b ()b (July/Summer)) (July/Summer).. 50 50
40 OperatingOperating pre pre lockdown lockdown conditions conditions were were characterized characteri40 zed by aby peak a peak load load of about of about 60 GW 60 inGW summer in [GW] (58.81[GW] 30summer GW on(58.81 25 JulyGW 2019)on 25 and July a 2019) minimum and a load minimum of 2030 GW load in of winter 20 GW (18.27 in winter GW on (18.27 25 December GW on 2019).25
In winter, on the contrary, the peak load is 55 GW andLoad the summer minimum load is 21 GW [9,14]. Load December 2019). In winter, on the contrary, the peak load is 55 GW and the summer minimum load 20 20 is 21 GW [9,14]. 3.10 Impact on Electricity Consumption and Load Profiles10 during COVID-19 Working Day Winter Holyday Winter Working Day Summer Holyday Summer 03. ImpactThe lockdown on Electricity sequence Consumption consists of and three Load main Profiles dates:0 during COVID-19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 The lockdown sequenceTime [h] consists of three main dates: Time [h] 6th March 2020: closing of all schools and universities; • • 12th6th MarchMarch 2020:2020: closing Alla) territory of all schools of Italy and becomes universities a “red; zone”. Non-essentialb) factories, companies • • and12th publicMarch buildings 2020: All areterritory closed of (partial Italy becomes shutdown); a “red zone”. Non-essential factories, companies Figure 3. Load profile (a) Weekday/Weekend (January/Winter) (b) (July/Summer). 21stand March public 2020: buildings airports are closed and seaports (partial closedshutdown); to passenger traffic (shutdown). • OperatingThe forced pre closure lockdown of public conditions buildings, were industries characteri and tertiaryzed by activities a peak duringload of COVID-19 about 60 outbreakGW in summerrestrictions (58.81 highly GW reduced on 25 theJuly consumption 2019) and a curves minimum during load the of daily 20 workingGW in winter hours and(18.27 in theGW evening. on 25 December 2019). In winter, on the contrary, the peak load is 55 GW and the summer minimum load is 21 GW [9,14].
3. Impact on Electricity Consumption and Load Profiles during COVID-19 The lockdown sequence consists of three main dates: • 6th March 2020: closing of all schools and universities; • 12th March 2020: All territory of Italy becomes a “red zone”. Non-essential factories, companies and public buildings are closed (partial shutdown);
Energies 2020, 13, x FOR PEER REVIEW 5 of 22
• 21st March 2020: airports and seaports closed to passenger traffic (shutdown).
EnergiesThe2020 ,forced13, 3357 closure of public buildings, industries and tertiary activities during COVID-195 of 19 outbreak restrictions highly reduced the consumption curves during the daily working hours and in the evening. In Figure 4, the daily load profile of the days 6th and 21st (for 2020, 23rd, the first Inavailable Figure4 ,Monday) the daily loadMarch profile is compared of the days for 6th the and last 21st three (for years, 2020, 23rd,although the firstneglecting available weather Monday) Marchdifferences is compared and further for the related last three economic years, issues, although a consistent neglecting reduction weather of di thefferences consumption and further has been related economicregistered. issues, As a result a consistent of the lockdown, reduction the of thedaily consumption consumption hascurve been changed registered. their typical As a result pattern. of the lockdown,In the morning, the daily electricity consumption consumption curve changedramps up their more typical slowly pattern. and the In evening the morning, peaks remain electricity consumptionsustained as rampspeople up prepare more slowlymeals and and spend the evening time with peaks their remain families sustained watching as peopleTV, using prepare electric meals andappliances spend timeand withelectronic their devices; families these watching peaks TV, are using anyway electric lower appliances than one andwould electronic normally devices; be theseexpected. peaks are anyway lower than one would normally be expected.
FigureFigure 4. 4. ItalianItalian Load profile profile comparison comparison 6th 6th March March (a ()a and) and 20–21 20–21 March March (b) ( b2018,) 2018, 2019, 2019, 2020. 2020. The evening peak is reached with a much greater gradient than in the past, putting under stress The evening peak is reached with a much greater gradient than in the past, putting under stress the regulation capabilities and flexibility of thermoelectric power plants that are working close to the the regulation capabilities and flexibility of thermoelectric power plants that are working close to the minimum, especially if the regulating capacity of hydroelectric plants (impoundment facility and minimum, especially if the regulating capacity of hydroelectric plants (impoundment facility and pumpingpumping storage) storage) shouldshould not be sufficient. sufficient. However, However, the the electric electric system system must must be beable able to manage to manage at at anyany time time such such changeschanges inin loadload in a range of of values values that, that, in in 2018, 2018, fluctuated fluctuated between between a minimum a minimum of of 20.220.2 GW GW and and a a maximummaximum of 57.8 GW. GW. To To assess assess that that issue issue the the residual residual load, load, defined defined as the as thedifference difference betweenbetween current current load load and and production production fromfrom renewablerenewable sources, has assumed remarkable remarkable importance importance in thein lastthe last years, years, due due to to the the increase increase of of the the renewable renewable resources, in in particular particular with with the the high high penetration penetration ofof PV PV plants. plants. TheThe residualresidual load corresponds to to the the actual actual load load that that must must be becovered covered by bydispatchable dispatchable generationgeneration plants.plants. Besides,Besides, the shape of of the the residual residual load load has has evolved evolved in inthe the last last years, years, even even more more didifferentiatingfferentiating from from the the pattern pattern of theof wholethe whole electrical elec profile,trical profile, and this and phenomenon this phenomenon will be increasinglywill be evidentincreasingly in the evident forecasted in the future forecasted scenarios future due scenarios to the expecteddue to the increase expected in increase PV generation. in PV generation. In the days ofIn this the observation, days of this theobservation, residual loadthe residual curve is load characterized curve is characterized by the typical by duckthe typical curve, duck with curve, relevant variationswith relevant during variations daytime during and in thedaytime steepness and ofin thethe eveningsteepness ramp of the due evening to the simultaneous ramp due to increase the insimultaneous demand and increase reduction in demand in renewable and reduction energy in production, renewable energy which determinesproduction, which the need determines for a rapid increasethe need in for production a rapid increase from programmable in production from sources. programmable The RES growth, sources. occurred The RES in growth, the Italian occurred territory, hasin the shown Italian to territory, be often has not shown consistent to be withoftenthe not pointsconsistent of majorwith the consumption. points of major On consumption. the other hand, On the other hand, especially for PV, the installation of the plants led to increasing network especially for PV, the installation of the plants led to increasing network congestion situations, due to congestion situations, due to the asymmetry in their installation location (most of the production is the asymmetry in their installation location (most of the production is located in South Italy for the located in South Italy for the solar energy availability). solar energy availability). In Figure 5 the variation of the peak in the demand is considered for March 2020 and the first In Figure5 the variation of the peak in the demand is considered for March 2020 and the first week of April. The profiles show how increasing lockdown measures caused a sustained decrease in weeknational of April. energy The demand. profiles In showparticular how, from increasing the last lockdown weeks of March, measures a decrease caused in a load sustained was observed, decrease in nationalfollowing energy the consecutive demand. In closure particular, of production from the facilities. last weeks In ofFigure March, 6, the a decrease peak variation in load iswas compared observed, followingfor the last the three consecutive years. To closurebetter clarify of production the phenom facilities.ena, in Figure In Figure 7 is 6reported, the peak a comparison variation is between compared for2018, the last2019 three and years.2020 Toof betterthe peak clarify variation the phenomena, profiles. Figure in Figure 7 7shows is reported an evident a comparison reduction between in 2018,consumption, 2019 and 2020which of reached the peak at variation least 25% profiles. less consumption Figure7 shows than the an evidentaverage reductionin the previous in consumption, year. which reached at least 25% less consumption than the average in the previous year. The analysis was brought to a higher level of detail, to evaluate the impacts of the health emergency, on all the areas of Italy. In particular, the profiles on the northern part of Italy was evaluated. The northern part of Italy, composed by Valle d’Aosta, Piemonte, Liguria, Emilia-Romagna, Lombardia, Veneto, Friuli Venezia Giulia and Trentino Alto Adige is the main productive zone of the nation, accounting for about half of the national gross domestic product. Figure8 shows the load profiles Energies 2020, 13, 3357 6 of 19 in the North of Italy. As shown in Figure8, their trends reflect those of the whole nation, describing a lowering of the load by almost half compared to normal conditions. The northern part of Italy is characterized by the highest penetration of industrial facilities and the impact on consumption of the lockdown was quite remarkable. Energies 2020, 13, x FOR PEER REVIEW 6 of 22
Pre - lockdown First Days lockdown Partial lockdown Full lockdown Full lockdown th st 1st – 7th March 8th – 14th March 15th – 21st March 22nd – 29th March 30th March – 5th April 47,09 GW 45,30 GW 47,09 GW 45,30 GW 40,74 GW 37,24 GW 37,60 GW @ 4th March @ 11th March th th st @ 4 March @ 11 March @ 18th March @ 25th March @ 1st April
-4,3% -15,1% -22,4%
FigureFigure 5. 5.Five-week-load Five-week-load profile profile,, 1st 1st March March–5th–5th April April 2020 2020..
FigureFigure 6. Five-week-load6. Five-week-loadprofile profile comparison between between 2018, 2018, 2019 2019 and and 2020. 2020.
Figure 7. Five-week-peak comparison between 2018, 2019 and 2020.
Figure 7. Five-week-peak comparison between 2018, 2019 and 2020.
1
Energies 2020, 13, x FOR PEER REVIEW 7 of 22
The analysis was brought to a higher level of detail, to evaluate the impacts of the health emergency, on all the areas of Italy. In particular, the profiles on the northern part of Italy was evaluated. The northern part of Italy, composed by Valle d’Aosta, Piemonte, Liguria, Emilia- Romagna, Lombardia, Veneto, Friuli Venezia Giulia and Trentino Alto Adige is the main productive zone of the nation, accounting for about half of the national gross domestic product. Figure 8 shows the load profiles in the North of Italy. As shown in figure 8, their trends reflect those of the whole nation, describing a lowering of the load by almost half compared to normal conditions. The northern
Energiespart of2020 Italy, 13 ,is 3357 characterized by the highest penetration of industrial facilities and the impact on7 of 19 consumption of the lockdown was quite remarkable.
Figure 8. 8. NorthNorth Italy Italy load load profile profile comparison comparison (a)( 6tha) 6th March March and and (b) (21b) March 21 March 2018, 2018, 2019 2019 and 2020. and 2020. 4. Impact on Electricity Market during COVID-19 4. Impact on Electricity Market during COVID-19 4.1. Italian Electricity Market 4.1. Italian Electricity Market In Italy, the electricity market (Italian Power Exchange—IPEX) was set up as a result of Legislative In Italy, the electricity market (Italian Power Exchange—IPEX) was set up as a result of Decree no. 79 dated 16 March 1999 (“Bersani Decree”) as part of the implementation of the EU directive Legislative Decree no. 79 dated 16 March 1999 (“Bersani Decree”) as part of the implementation of onthe the EU creation directive of on an internalthe creation energy of marketan internal (Directive energy 96 market/92/EC (Directive repealed by96/92/EC Directive repealed 2003/54 by/EC). DirectiveThe Italian2003/54/EC). electricity market is divided into: - TheDay-ahead Italian electricity market (MGP); market is divided into: -- Day-aheadIntra-day market market (MI);(MGP); -- Intra-dayDispatching market services (MI); market (MSD). - Dispatching services market (MSD). In the MGP and MI—also referred to as energy markets—producers, wholesalers and end In the MGP and MI—also referred to as energy markets—producers, wholesalers and end customers, as well as Acquirente Unico (AU), responsible for acquiring energy for captive customers customers, as well as Acquirente Unico (AU), responsible for acquiring energy for captive customers and Gestore dei Servizi Energetici (GSE), responsible for selling energy for renewable producers, and Gestore dei Servizi Energetici (GSE), responsible for selling energy for renewable producers, buy buy and sell wholesale quantities of electricity for the next day. These markets, which are managed by and sell wholesale quantities of electricity for the next day. These markets, which are managed by GestoreGestore dei Mercati Mercati Energetici Energetici (GME), (GME), define define system system marginal marginal prices prices at which at which the energy the energy is traded. is traded. InIn thethe MSD,MSD, the the transmission transmission system system operator operator (TSO), (TSO), Terna, Terna, procures procures the the resources resources it needs it needs to manage to andmanage control and the control system the (solving system (solving intranational intranationa congestions,l congestions, creating creating energy energy reserves, reserves, real-time real-time balancing). Givenbalancing). that transportGiven that capacity transport limits capacity exist, limits the Italian exist, the power Italian system power is subdividedsystem is subdivided into market into zones andmarket are zones cleared and through are cleared the localthrough market the local algorithm. market algorithm. The number The of number constrained of constrained zones or zones “point of limitedor “point production”, of limited production”, are zones whose are zones maximum whose maximum generation generation exportable exportable to the rest to of the the rest grid of isthe lower thangrid is the lower maximum than the possible maximum generation possible generation owing to owing insuffi tocient insufficient Transmission Transmission capacity capacity [14]. Currently, [14]. theCurrently, zones are the aszones follows are as (Figure follows9): (Figure 9): • Geographical zone: zone: representing representing a aportion portion of ofthe the national national grid. grid. Geographical Geographical zones zones are northern are northern • ItalyItaly (NORD),(NORD), central-northerncentral-northern ItalyItaly (CNOR), (CNOR), central-southern central-southern Italy Italy (CSUD), (CSUD), southern southern Italy Italy (SUD), Sicilia(SUD),(SICI), Sicilia (SICI), Sardegna Sardegna (SARD); (SARD); • National virtual zone: constrained zone (“point or pole of limited production”). It includes National virtual zone: constrained zone (“point or pole of limited production”). It includes • Rossano (ROSN); Rossano (ROSN);
Foreign virtual zone: point of interconnection with neighboring countries. It includes France • (FRAN), Switzerland (SVIZ), Austria (AUST), Slovenia (SLOV), Slovenia coupling representing the interconnection dedicated to the market coupling between Italy and Slovenia (BSP); Corsica (CORS), Corsica AC (COAC), Greece (GREC), France coupling (XFRA), Austria coupling (XAUS), Malta (MALT), Switzerland coupling (XSVI) and Montenegro (MONT). Market zone: aggregation of geographical and/or virtual zones such that the flows between the • same zones are lower than the transmission limits notified by the TSO. This aggregation is defined on an hourly basis as a result of the resolution of the day-ahead market and intra-day market. In the same hour, different market zones may have different zonal prices. Energies 2020, 13, x FOR PEER REVIEW 8 of 22
• Foreign virtual zone: point of interconnection with neighboring countries. It includes France (FRAN), Switzerland (SVIZ), Austria (AUST), Slovenia (SLOV), Slovenia coupling representing the interconnection dedicated to the market coupling between Italy and Slovenia (BSP); Corsica (CORS), Corsica AC (COAC), Greece (GREC), France coupling (XFRA), Austria coupling (XAUS), Malta (MALT), Switzerland coupling (XSVI) and Montenegro (MONT). • Market zone: aggregation of geographical and/or virtual zones such that the flows between the same zones are lower than the transmission limits notified by the TSO. This aggregation is defined on an hourly basis as a result of the resolution of the day-ahead market and intra-day Energies 2020, 13, 3357 8 of 19 market. In the same hour, different market zones may have different zonal prices.
FigureFigure 9. 9.Italian Italian load load subdivision subdivision zones zones for for the the Italian Italian Power Power Exchange Exchange (IPEX). (IPEX). The day-ahead market algorithm runs the first time at the national level. If the capacity limits The day-ahead market algorithm runs the first time at the national level. If the capacity limits among zones are not reached, the outcome is an equal price for all the zones, otherwise, the market is among zones are not reached, the outcome is an equal price for all the zones, otherwise, the market split into one or more zones. All supply offers and demand bids are valued at the marginal clearing is split into one or more zones. All supply offers and demand bids are valued at the marginal clearing price of the zone to which they belong. This price is determined, for each hour, by the intersection of price of the zone to which they belong. This price is determined, for each hour, by the intersection of the zonal demand and supply curves and is different from one zone to another zone when transmission the zonal demand and supply curves and is different from one zone to another zone when capacity limits are saturated. The accepted demand bids pertaining to consuming units belonging to transmission capacity limits are saturated. The accepted demand bids pertaining to consuming units Italian geographical zones are valued at the “Prezzo Unico Nazionale” (PUN—national single price); belonging to Italian geographical zones are valued at the “Prezzo Unico Nazionale” (PUN—national this price is the weighted mean of the prices of the geographical zones, weighted for the quantities single price); this price is the weighted mean of the prices of the geographical zones, weighted for the purchased in these zones [14]. quantities purchased in these zones [14]. Commonly, Italy shows significant differences among the market prices of different zones. This is Commonly, Italy shows significant differences among the market prices of different zones. This due to bottlenecks in the north-south transmission lines that are unable to manage the regional is due to bottlenecks in the north-south transmission lines that are unable to manage the regional imbalance of generation and demand between the two zones. Currently, plans have been made to imbalance of generation and demand between the two zones. Currently, plans have been made to strengthen transmission lines and to achieve further transmission capacity to accommodate higher strengthen transmission lines and to achieve further transmission capacity to accommodate higher renewables according to the national electricity transmission grid development plan [12] and ENTSO-E renewables according to the national electricity transmission grid development plan [12] and TSO cooperation regarding the network development and long-term planning defined in Ten-Year ENTSO-E TSO cooperation regarding the network development and long-term planning defined in Network Development Plan (TYNDP) [16]. For instance, the construction of three main high voltage Ten-Year Network Development Plan (TYNDP) [16]. For instance, the construction of three main high direct current (HVDC) lines is being considered, which will connect south and north of Italy through voltage direct current (HVDC) lines is being considered, which will connect south and north of Italy the mainEnergies Islands 2020 of, 13 Sardinia, x FOR PEER and REVIEW Sicily (Figure 10)[17]. 9 of 22 through the main Islands of Sardinia and Sicily (Figure 10) [17].
HV DC SA.CO.I.3
HV DC SA.PE.I
HV DC Tyrrhenianlink
Figure 10.FigureThe 10. option The option of building of building a new a new HVDC HVDC Tyrrhenian Tyrrhenian link link to to Sicily Sicily/Sardinia/Sardinia—the—the year 2030 2030..
Besides, the growth of renewable power has forced down wholesale electricity prices given that the clearing power price is determined by the “merit order”—the sequence in which power producers contribute power to the market, with the cheapest offer made by the producer. Power from renewable installations such as wind and PV is offered at null cost (price takers). That means they lower the entrance price and push more expensive conventional producers down in the merit order list (Figure 11).
Pr i ce (€/ M Wh) Peaking Peaking gener ati on
Turbogas Oil gener ati on Inter m ed i ate Base gener ati on
Gas Coal Renew abl e gener ati on Import Wind, Solar, H ydro
Quantity(MWh)
Figure 11. Merit order sequence for bidding in the IPEX market.
The growth of renewable sources has led to a significant decrease in the hours of use of thermal plants with the highest production costs and a significant reduction in the PUN. As can be seen in Figure 12, high quantities of RES lead to low prices of energy in the market, specially recorded in the summer months.
Energies 2020, 13, x FOR PEER REVIEW 9 of 22
Energies 2020 13 , , 3357 9 of 19 Figure 10. The option of building a new HVDC Tyrrhenian link to Sicily/Sardinia—the year 2030. Besides, the growth of renewable power has forced down wholesale electricity prices given thatBesides, the clearing the growth power of price renewable is determined power has by theforced “merit down order”—the wholesale electricity sequence prices in which given power that producersthe clearing contribute power price power is determined to the market, by the with “merit the cheapest order”—the offer sequence made by in the which producer. power Power producers from contributerenewable power installations to the market, such as with wind the and cheapest PV is o offerffered made at null by the cost producer. (price takers). Power That from means renewable they installationslower the entrance such as price wind and and push PV more is offered expensive at nu conventionalll cost (price producers takers). That down means in the they merit lower order the list (Figureentrance 11 price). and push more expensive conventional producers down in the merit order list (Figure 11).
Price (€/ MWh) Peaking Peaking generation
Turbogas Oil generation Intermediate Base generation Gas Coal Renewable generation Im port Wind, Solar, Hydro
Quantity(MWh)
Figure 11. Merit order sequence for bi biddingdding in the IPEX market.
The growth of renewable sources has led to a significant significant decrease in the hours of use of thermal plants with the highest production costs and a sign significantificant reduction in the PUN. As can be seen in Figure 12,12, high quantities of RES lead to low prices of energy in the market, specially recorded in the summerEnergies 2020 months., 13, x FOR PEER REVIEW 10 of 22
Price Price (€/ MWh) (€/ MWh) High High High High Demand Demand Low Low Demand Demand Low Renewable
4 3 2 1 T.gas T.gas Renewable Im port Coal Gas Oil Renewable Im port Coal Gas Oil Quantity(MWh) Quantity(MWh) a) b)
Figure 12. (a) dashed line 2: price with low demand/low RES, dashed line 4 price with high Figure 12. (a) dashed line 2: price with low demand/low RES, dashed line 4 price with high demanddemand/low/low RES, ( (bb)) dashed dashed line line 1: 1: price price with with low low demand/h demandigh/high RES RES and and dashed dashed line line4: price 4: price withwith highhigh demanddemand/high/high RES. RES. 4.2. Statistical Comparison of Pre- and Post-Lockdown Day-ahead Market Prices 4.2. Statistical Comparison of Pre- and Post-Lockdown Day-Ahead Market Prices To understand the effects of the COVID-19 lockdown, a statistical analysis was performed, To understand the effects of the COVID-19 lockdown, a statistical analysis was performed, aimingaiming toto comparecompare the different different pre pre and and post post lockdown lockdown MGP MGP prices prices distributions. distributions. FiguresFigures 13 and 1414 showshow the the boxplots boxplots of of the the rela relativetive deviation deviation from from the themean mean of pre-lockdown of pre-lockdown WD P marketmarket pricesprices pre, ,in in order order to to perform perform a diff-in-diff a diff-in-di analysis.ff analysis. ℎ : the observed hourly PUN prices ( ) were aggregated in sets of observed prices per each week of the reference year, for days between Monday and Friday (weekdays), represented by the sets { }. Then, sets of their relative differences (in percentage) from the pre-lockdown mean price { } were computed using (1), where the pre-lockdown mean price was calculated as in (2).