break down of the former vertically integrated monopolist (ENEL) to be realized through auction sales to the private market of three GENCO (new companies ...
ELECTRICITY DEMAND IN WHOLESALE ITALIAN MARKET
Simona Bigerna*
*Department of Economics, Finance and Statistics - University of Perugia (I) Via Pascoli 20, 06123, Perugia
ABSTRACT
In this paper I pursue two objectives: the first is to construct a model behavior of electricity demand in the IPEx, the second is to measure demand elasticity at hourly level, directly from consumer behavior. In the literature, demand elasticity estimation has been done constructing a residual demand curve, using the supply or offer bid data, in order to measure unilateral market power through the Lerner index. This paper takes a novel approach, providing the first attempt in the literature to estimate demand elasticity using demand data, in the IPEx, using duality approach to derive a hourly demand function for electricity. Econometric estimation allows to ascertain that elasticity varies significantly with time of the day, with day of the week, with the existence of line congestion and according to the level of the equilibrium price. Another interesting issue is how elasticity varies along the demand curve. Estimation results show that elasticity is generally higher and tends to increase in the portion below the actual equilibrium price. This means that more competition on the supply side can yield lower equilibrium prices and proportionately much higher quantities, because a lower offer curve, shifted to the right would intersect a flatter portion of the demand curve.
Key words: electricity demand; hourly data; elasticity JEL-Code: D12; L10; Q41.
INTRODUCTION Estimation of demand elasticity has been, ever since the foundation of modern economic theory, one of the most popular empirical exercises in applied economics. The relevance of knowing demand elasticity is quite obvious for advancement of research about consumer preferences and consumer behaviour, as well as for guidance to the adoption of policy measure, ranging from taxation to welfare. For instance, taxonomy of elastic vs. inelastic price response commodities is crucially based upon quantitative measure of demand elasticity. In general, the typical data used for empirical estimation is market prices and quantities. This means that we observe one point along the demand curve in every state of the world realisation and identification of a (stable) demand function can be obtained using appropriate covariates in econometric estimation procedure. This is true whether one uses aggregate market data, pretending to infer a representative consumer’s behaviour or one uses individual disaggregate data, coupling observed quantities purchased with market prices (Ref. ). Very seldom we find in the literature estimation based on data reflecting ex ante consumers willingness to pay, as it is stated in every Economics text book, for a whole range of price and quantity pairs. As a paradox in applied consumer analysis, most of estimation using ex ante observed willingness to pay data are based on survey data, focusing on public goods and externalities, such as environmental quality, air pollution, social services rendered by local municipalities, public transportation and so on. So it is not unfair to say that, essentially, present state of the art in empirical demand analysis is focused along two alternative and partially unsatisfactory lines: on the one hand, there is estimation of individual preferences based on market data; on the other hand, there is survey of consumer intention to purchase “goods” for which there does not exist a market. In this paper, I would like to present a novel approach for estimation of demand elasticity, using collected and well organised data for ex ante consumer demand schedule, based on a large data set of consumer price and quantity bids for electricity in an organized market. Notably, I use demand bids data in the day-ahead market in the Italian Power Exchange (IPEX). In this market, individual bidders on the demand side (for short, from now on: “consumers”) send to market organiser a bid consisting of a pair of price and quantity for each hour of the following day; the quantity is the expression of the amount of KWh that consumer wants to purchase (and is kept liable to consume); the price is the expression of the unit value of KWh that consumer intends to pay for that quantity. In the day-ahead market, consumer bids simultaneously price and quantity pairs for 24 hours, thus expressing a well defined willingness to pay for each quantity, presumably according to a complete and well structural behavioural strategy. Surprisingly, these data have never been used before to estimate demand elasticity, while many empirical analyses of electricity market have used producer bids for price and 2
quantity in the day-ahead market (Wolak 2003b and 2010) and some have estimated demand elasticity from market equilibrium data. These studies typically assume a specific market structure, i.e. oligopolistic behaviour and then compute residual demand for each market participant from which some parameter inference about (residual) demand elasticity can be recovered. Therefore it is somehow surprising that “demand” behaviour is estimated from “supply” data, while there is a more direct and obviously way to estimate “demand” behaviour, i.e. using demand bid data. My critique to the above mentioned approach is that the notion of demand is the notion of “residual demand” facing an oligopolistic supplier which is obtained subtracting from total market ex post realized equilibrium quantity the ex ante quantity bid by individual oligopolist. There are two problems with this approach. The first problem is that each supplier bids in the ex ante day ahead market, without knowing total market demand quantity; each supplier can, at best, make a efficient forecast about “expected” market demand. The second problem is that computation of “residual demand” is conditional to the maintained hypothesis that suppliers behave as Cournot oligopolists. It is evident from the above considerations that individual and/or aggregate consumer behaviour information is not used at all. So, there is certainly a loss of efficiency in using only suppliers bid data, plus a potential mis-specification issue in case of other-thanCournot behaviour in the supply side. Some studies have used electricity market equilibrium data to estimate demand elasticity. An excellent survey of recent empirical work can be found in Estimation of elasticity price of electricity with incomplete information Xavier Labandeira, José M. Labeaga, Xiral López-Otero, Energy Economics, 2011, other recent studies are: Regional Differences in thePrice-Elasticity of Demand For Energy Mark A. Bernstein, James Griffin, Prepared for the National Renewable Energy Laboratory Rand Technical Report 2005, Applied Economics, Consumer welfare effects of increased food and energy prices Kuo S. Huanga; Sophia Wu Huang, US Department of Agriculture, Economic Research Service, Washington, DC, USA First published on: 03 May 2011. Not surprisingly, using this data is not uncontroversial, as stated by Fezzi and Bunn (2010): ”Thus, there remains a prominent interest in modelling the interactions of supply and demand as a joint system, in particular to investigate the elusive demand elasticity issue from the perspective of publicly available, wholesale day-ahead market data. (OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 72, 6 (2010) 0305-9049 doi: 10.1111/j.14680084.2010.00596. Structural Analysis of Electricity Demand and Supply Interactions). Some other studies have used individual household information, collected from billing data. Again, these data refer to realized purchases by individual household and not from ex ante demand function. (for instance, Reiss White 2008 use billing data from san Diego area; other studies are: Reiss White 2005 Estimating demand with non linear prices) . Another relevant problem highlighted in the previous literature Is that “Electricity demand elasticities are a subject of nearly endless contention. The relevant elasticity would be a short-run elasticity in the sense of the customer’s ability to respond to potentially large 3
hourly price volatility, but still recognizing that customers would know well in advance that prices could be quite volatile. The actual elasticity will depend in great part on technology, as automated response to price changes will surely become easier”, thus longrun elasticities might be larger (Borenstein CSEM WP 133, The Long-Run Effects of Real-Time Electricity Pricing, Severin Borenstein, June 2004). In order to identify a well defined demand functions in the electricity market, it is wise to notice that demand data are expressed by individual consumers in the IPEX in two different ways, distinction being crucially relevant for present analysis. Some consumers express a quantity bid without corresponding indication of a price they would be willing to pay: these are consumers who show a perfect inelastic behaviour, as they are in principle willing to pay any price that would result from market clearing procedure. The IPEX market clearing procedure assigns a default price limit to these bids which has been varying in time, bat that heuristically reflects the maximum price cap to suppliers that is allowed by the Regulatory Authority within the market mitigation and competition policy measures. As a practical example in the last year (2010), to these bids there has been assigned a default price of 3000 euro/MWh. Some other consumers express a simultaneous quantity and price bid. These are obviously somehow elastic consumers, for they express a willingness to purchase a certain electricity quantity only if they can pay a certain well defined price. Obviously, when all individual behaviours are aggregated in a “market demand function”, the shape of the demand schedule is a vertically down ward sloping curve until the portion of “elastic consumers” is reached, from this point onward demand schedule shows a well defined negative slope. In principle, there are two possible market clearing outcomes, which are crucially relevant here. One possible market outcome is determined by intersection of upward sloping supply curve with downward sloping portion of demand curve (see fig.1.A); another possible market outcome occurs when the upward supply curve intersects vertical portion of demand curves (Figure 1.B).
4
FIGURE 1 Possible market clearing outcomes in the electricity market Figure 1A
Figure 1B
Domnda - Offerta (Ore12)
Domnda - Offerta (Ore12)
3000
3000 D1 S1
2500
2500
2000
2000
1500
1500
P
P
D1 S1
1000
1000
500
500
0
0
2
4
6 Q
8
10
12
0
0
4
x 10
2
4
6 Q
8
10
12 4
x 10
It is clear from these considerations that the concept of market demand elasticity is well defined only in the first case. It is therefore an empirical issue to classify market outcomes according to observed demand elasticity in order to make realistic conjectures about consumer behaviour. The setup of this paper is as follows. Section describes the complexity of the IPEX due, primarily, to the geographical segmentation. Section three we describes the theoretical framework. Section 4 describes methodology, data sources and the build of our large data bank. Empirical results are shown in section 5, while section 6 concludes and summarizes main findings.
1. THE IPEX NORMATIVE FRAMEWORK
There are some features of the Italian electricity market which are relevant to understand the operational way in which I have modeled the consumer behavior for practical estimation. The Italian liberalization process began under the framework of the European Directive of energy sector liberalization, 96/92/CE with national legislation (Law 79/99) enacting a break down of the former vertically integrated monopolist (ENEL) to be realized through auction sales to the private market of three GENCO (new companies controlling about 50% of the existing generation capacity. In addition, there has been created by other legislation (Law 07/02 and 240/04) an Independent System Operator (now: TERNA), a Market Operator, (GME) managing a non-compulsory pool market and a a Single Buyer (AU) in charge of aggregating small and poor customer demand.
5
Similar to other countries (Newbery, 2005), the Italian market has been organized as a sequence of three markets: day ahead, adjustment and dispatching resource market (“Ancillary services market”). The equilibrium price in the day ahead market is set as the system marginal price (SMP) for each zone in which the market is separated due to network congestion, as measured by excess of physical transmission capacity, based on supply and demand bids (Bollino Polinori 2007). These latter are the prices received by the supply side, while on the demand side there is a unique national price, set as a weighted average of zonal supply prices.
The adjustment market allows generators and loads to correct parts of schedules which cannot be implemented due to technical constraints. Ancillary service market is a single market allowing the System Operator to procure congestion relieve and reserve margin resources, on a pay-as-bid basis. Other main characteristics of the IPEX have been: (i) in the period April – December 2004 only suppliers were allowed to participate into the market, while demand was, inelastically, represented by the best day-ahead forecast of System Operator; (ii) in January 2005 active demand bids entered in to the market, while the System Operator maintained the privilege to bid in case of endangering of system reliability (i.e.,v when total market demand is “too different” from day-ahead forecast used for network security management; (iii) in January 2005 the AU was instructed to use contract for differences extensively and buy into the market. As a result, market liquidity rose to above 60%, due to AU dimension. (iv) in 2007 there has been a legislative attempt to reform the market, aimed at discouraging producer quantity withholding strategies, essentially “threatening” those generators found liable of market power abuse to be paid on a “pay-as-bid” basis, rather than receiving the “system marginal price”. energy payment to a supplier, who withholds quantity aimed at exercising market power. Recently, this provision was repealed due to a Court decision. (v) in 2009 a new reform was announced, trying to change the price setting from the system marginal price to pay as bid, but the operational implementation is still pending.
2. THEORETICAL FRAMEWORK
In the Italian wholesale electricity market there are represented distinct groups of consumers, such as large consumers (energy intensive industries railways, telecom companies), industrial consumers, traders who typically intermediate residential consumers (households) In principle, residential consumers demand electricity in order to produce other flows of relevant services (such as heating, cooking and so on), to maximize utility, while
6
industrial consumers use electricity as an input in the production function to produce goods and services, within a cost minimization process. Using duality approach, in this paper I assume that for both residential and industrial consumers it is possible to postulate the existence of a cost function for using electricity as a good “e” and a composite numerary good “y”: (1) c = c (pe,py, x) where pe is the price of electricity, py is the price of the composite numerary good and x id the objective variable (utility for residential consumers, output for industrial consumers). The easiest way to make use of the theoretical specification of function (1) is to consider “e” as the total amount of electricity used by the consumer in a given period. Considering that IPEX is a hourly market, I can refer equation (1) to the hourly cost function, from which it is straightforward to derive a hourly demand function for electricity, using Shephard’s Lemma (time subscripts are omitted for simplicity): (2) e = ∂c/∂pe = e (pe,py, x) Demand function (2) expresses quantity demand for electricity as a function of own price, other goods price and the objective variable (typically, output). While it is formally legitimate, it could be unpractical, if not even unrealistic, to consider a hourly demand for electricity depending, among other things, from a hourly price of all other goods. As stated before, use of duality allows to recover also Marshallian demand functions. It suffices to invert eq (1) into a inverse utility/production function, to obtain: (3) x = x(pe, py, c) and then apply Roy’s Identity to eq (3), in order obtain: (4) e = - ∂x/∂pe / ∂x/∂c = e (pe, py, c) where, as before, in (4) pe is the aggregate price of electricity, py is the price of the composite good and c is the total consumer expenditure. Equation (4) holds for each state of nature of demand. Typically, it is assumed to hold for each hour of an hourly market as it is the Italian market. The attractiveness of equations (4) is that it is quite straight forward to estimate and compute the elasticity of demand, with respect of both price and total expenditure. i.e price elasticity and income elasticity.
7
3. DATA AND ECONOMETRIC MODEL 3.1 Data The data used are made available by the Market operator in electronic format. In this paper I use data from jan 2005 to sept 2011, discarding year 2004 data when demand bids was not operational. Each month amounts to about 1 million records, reporting, type of market, bid quantity and price, status of accepted or rejected, awarded prices, accepted quantity, status of marginal plant, zone, plant and producer ID, trader ID. There are seven geographical zones in Italy, which can be aggregated in homogeneous segmented zones according to the existence of network congestion. I define a homogeneous segmented zone in each hour a zone cluster characterized by the same SMP (NOTE: I exclude few cases of virtual zones which are defined as a single plant with limited production for grid security purposes). I used STATA and TSP programs to select homogenous segmented zones and system marginal price (SMP), to identify all participants to each segmented zone, to compute aggregate demand for each segmented zone. In the Italian market there were defined by the Energy Authority 4 different hourly clusters (NOTE: Hourly clusters are labeled F1 F2 F3 F4 and are roughly representing, respectively, very high peak load hours, daily peak load hours, weekdays off-peak load hours, night and weekend hours. This clustering has been used to perform market monitoring and surveillance by the Energy Authority. It has been modified and the dismissed in 2008). Market participants (and the Authority alike) expectations have usually been that in during peak (off-peak) hour period prices had necessarily to be higher (lower) than average. Thus, hourly clusters work like price signals, in the sense that producers embody the expectation that they could be somehow legitimated to charge higher prices in peak hours. Market separation due to congestions occurs typically during peak hours. The number of segmented ones in each hour varies between one (no separation) and seven (each geographical zone constituting a segmented zone by itself). These latter extremes have very seldom occurred. The most common occurrences is that of two or three segmented zone markets (NOTE: Notice that this can imply different aggregations; for instance: “North and the Rest of Italy” and “Continental Italy+Sardinia and Sicily” are both feasible realizations of a two-segmented zone market. Obviously, not all combination of elementary zone in segmented zone markets are feasible: due to Italian transmission network shape (relatively long and narrow), only contiguous zones may be aggregated univocally along the North-South direction). This implies that statistical market data for a given geographical zone result from aggregation of different market outcomes; for instance, the “average annual price in North” averaged out hours in which equilibrium price resulted from North market alone with hours in which equilibrium price determined by a larger market aggregation.
8
I analyze market data according to price quartiles, characteristics of maket operators, hourly clusters, geographical zone, market segmentation. Italian demand for electricity shows the typical humped shape during the day and has not changed dramatically during the years (Fig 4.1), with a long term average annual growth rate about 2% (except during the downturn period due to the global financial crisis of 2008-2009). On the demand side the number of operators has increased in the first years and is now quite stable. With the exception of the Acquirente Unico (the Italian Single Buyer: a Government owned company which is in charge of purchasing electricity on the wholesale market on behalf of non-eligible customers and poor households), no operator has a significant market share (table 4.1). During the period considered, the first five traders share has not exceed 20%. In Italy, electricity prices have been structurally higher than in the rest of Europe and have followed quite closely oil and gas price trends (table 4.2). This Is not surprising, given that the Italian generation capacity is highly skewed toward hydrocarbons, which cover almost three quarters of total generation. Notice that zonal prices are quite different from the national average, due to differences in plant technology, network configuration and possibly generators strategic behavior (Bollino Polinori 2010), see table 4.3). Market zones aggregation due to congestion reflects demand levels and network operating conditions. In the period considered, market zones pattern has been quite stable, however, with some evolution due to transmission network developments which have been aimed at relieving congestions. During daily hours, the typical situation is three-zone segmentation: Continental Italy, Sicily, Sardinia. In peak hours, it is often he case that Center-North and Center-South break up into separate zones. The analysis of relative frequency of market segmentation (table 4.4) shows that roughly half of the times the Italian electricity market is split into 3 or more different zone, while Italy as a whole market (no segmentation) occurs only 2% of the times. The Italian duration curves (prices and quantities) computed by market zones and hourly clusters confirm previous results (Bollino and Polinori, 2005, p. 8), showing that price dispersion appears to be quite uncorrelated to load levels.
3.2 Econometric model In order to estimate empirically the demand functions, there is need to specify a parametric functional form. Unfortunately, most studies (as an examples for all, see Labandeira et al 2011) depart from the theoretical structure and introduce a linear or loglinear demand function specification, which does not necessarily respect the theoretical restrictions of consumer theory, embodied in eqs. (3) and (4), namely, adding up, symmetry and homogeneity.
9
In this paper, I assume two different functional forms for consumer preferences: GAI (Bollino 1986) and LES (Klein-Rubin 1947) functions. The GAI (Generalised Almost Ideal Demand System) is derived from the original AI system due to Deaton and Mulellbauer (1980) and is characterized by the introduction of committed quantities, which constitute both a plausible theoretical assumption and a convenient empirical device for cross section estimation. The LES is a linear demand system with committed quantities. I start with the aggregate electricity demand given by eq (4). I assume for simplicity that py=1 and that total expenditure C can be proxied by a group of relevant socio-economic determinants dj : C ≈ Σ βj dj. The parametric form of eq (4) in the case of GAI can be written as: (5) e = γ + α ln pe + β Σ βj dj Where γ is the committed quantity parameter, α is the price coefficient and β is the total expenditure coefficient. Econometric estimation of eq (5) can be easily performed and price elasticity can be computed as: ε = (∂e/∂pe).(pe/e) = α / e. Estimation uses all observations referring to a definite and explicit price bid, thus excluding those observations for which the demand schedule is perfectly inelastic (i.e., those bids without a price). Alternatively, the parametric form of eq (4) in the case of LES can be written as: (6) e = γ + (α/pe ) ( Σ βj dj ) and price elasticity can be computed as: ε = (∂e/∂pe).(pe/e) = α / pe e.
4. EMPIRICAL RESULTS
From the analysis of the Italian electricity market data I observe (Fig 4.1, Tables 4.1 and 4.2) several facts. Firstly, the daily profile is bimodal with a peak in the morning hours (around 11:00 12:00) and, especially in the Summer, a second peak around 19:00 20:00. In the Winter the second peak is much less evident. Secondly, there exist some market concentration on the demand side. Apart from the bilateral contracts and the Single Buyer activity, which account together for about half of the total demand, there are the first four large operators which account for about 25-30% and a fringe of numerous buyers (around 70 – 90) which account for the remaining 20% or less. 10
Thirdly, we observe that average market prices are increasing during the years according to the general energy price trend of the period while the difference between minimum and maximum price is quite large, for min price range around 20-25 euro/MWh, while max price reached easly the 170- 200 euro/MWh range, with an exceptional value of 378 euro/MWh in july 2006. Fourthly, we observe that zonal market separation due to congestion (table 4.4) is quite a structural phenomenon; the most frequent event is a two-zone separation, occurring 4045% of the hours. Next, a three-zone separation occurs around 30% of the times, while no separation (typically, in off peak hours) occurs around 15-20% of the times. Fifthly, we observe that zonal prices (table 4.5) tend to be higher in the Islands, showing prices around 25% higher than Italian average, while prices in Southern regions have been higher than in North in the period 2004-2008, but lower afterwards also because of the massive increase of renewable sources in the South, which have dispatching priority.
The elasticity of demand is estimated form the theoretical model (5) for every hour in the period 2009-2011. The results of elasticity estimation show clearly two main results, which are new results for the Italian electricity market literature: (i) there exist a well defined and statistically robust value for the demand elasticity, which can be estimated on average around (in absolute value) 0.03 - 0.10; (ii) elasticity values depend crucially on time of day and year and moreover from some characteristics of the market and of the zones. (Tables 4.6 to table 4.11). These Tables show the demand elasticity coefficients relative to different framework of analysis. Average monthly elasticities tend to be lower in 2011 than 2010. Considering demand elasticity aggregated by day of the week reveals that Sunday elasticities tend to be the lowest in the week (around -0.01 to – 0.05), while on Wednesday elasticity values tend to be highest (around -0.07 to -0.11). (table 4.7) Next, consider elasticity occurring in peak hours and off-peak hours. It is interesting to notice that the measure of elasticity varies significantly with time of the day. Demand elasticity aggregated by peak hours vs. off peak hours are quite different: during peak hours values are definitely larger, around -0.06 -0,08, while in off peak hours values are lower (table 4.8). Demand price elasticity values are different according to marginal operator determining the equilibrium price (table 4.9). Elasticity are differentiated as a result of the presence of congestion. (Table 4.10).The first line refers to a market characterized by the absence of congestion (National Single Market), while the second one shows the elasticity computed in the presence of two zones, that is when the market is split in two ones, and so on, up to five or more zones.
11
Demand price elasticity are different when there are different market splitting in zone division due to congestion showing a U shaped pattern: values are lower when there is less congestion (no zone division), -0,04 -0,05 and increase up to –0,07 -0,09 when there are 4 zones, but decrease wioth 5+ zone division. Demand price elasticity are different according to the absolute level of the equilibrium system marginal price (Table 4.12). In particular, values show a U shaped pattern, with the peak at the 7th-8th decile of the price distribution. Elasticity is also lower in the Summer in the last two years, possibly because consumers have acquired less elastic habits when the season requires air conditioning. Another interesting issue is how elasticity values can affect market equilibrium. We have already understood that estimation results show that elasticity is generally higher (in absolute value) when there is congestion, in peak hours, in the Winter and in the North. This means that in these situations more competition on the supply side can yield lower equilibrium prices and proportionately much higher quantities, because a lower offer curve, shifted to the right because of higher competition, would intersect a flatter portion of the demand curve.
5. CONCLUSION
In this paper I addressed the issue of analyzing consumer behavior in the new Italian deregulated electricity market. Data on market bids made available by the Energy Authority allowed us to compute demand elasticity for each trader on the demand side. Data are available are used over the period Jan 2005 - December 2011. The estimation of demand elasticity form demand bid data is a new result in the literature. The results of elasticity estimation for the Italian electricity market show clearly two main results, which are new results for the Italian electricity market literature; first, there exist a well defined and statistically robust value for the demand elasticity, which can be estimated on average around (in absolute value) 0.03 - 0.10; second, elasticity values depend crucially on time of day and year and moreover from some characteristics of the market and of the zones. Notably, estimated elasticities are generally higher (in absolute value) when there is congestion, in peak hours, in the Winter and in the North. This is important for competition fostering policies, because more supply competition resulting in a flatter offer curve can yield lower equilibrium prices, because the offer curve would intersect a flatter, more elastic, portion of the demand curve.
12
REFERENCES Baker, J.B. and , T.F. Bresnahan (1988). "Estimating the residual demand curve facing a single firm." International Journal of Industrial Organization 6 (3): 283-300. Bollino, C.A. and P. Polinori (2005). "Measuring Market Power in Wholesale Italian Electricity Markets: Preliminary results - April 2004." Paper prepared for 25th Annual North American Conference of the USAEE/IAEE, Denver, Colorado (USA) September 18-21, 2005. Borenstein, S., J. Bushnell and F. A. Wolak (2002). "Measuring Market Inefficiencies in California's Wholesale Electricity Industry." American Economic Review 92 (5): 13761405. GME (Italian Power Exchange), Monthly trading report April 2004, (Italian version only), http://www.mercatoelettrico.org/GmeWebInglese/default.aspx. GME (Italian Power Exchange), Monthly trading report December 2004, (Italian version only), http://www.mercatoelettrico.org/GmeWebInglese/default.aspx. Lerner, A. (1934). “The concept of monopoly and the Measurement of Market Power.” Review of Economics and Statistics 1 (June 1934): 157-75. Massey, P. (2000). “Market definition and market power in competition analysis: some practical issues.” The economic and social review 31(4): 309-28. Motta, M. (2004). “Competition policy. Theory and practice.” Cambridge University Press. Newbery, D. (2005). “Electricity liberalisation in Britain: The quest for a satisfactory wholesale market design.” The Energy Journal (special issue): 43-70. Ranci, P., A. Pototsching, E. Settimio, S., Frontini and A. Prandini (2005). “Review and potential of demand response measures.”, SESSA Conference: Investment for Sustainability, Madrid, 19-20 May. Sweeting, A. (2004). “Market Power in the England and Wales Wholesale Electricity Market 1995-2000.” Mimeo. Wolak, F.A. (2000). “An Empirical Analysis of the Impact of Hedge Contracts on Bidding Behaviour in a Competitive Electricity Market.” International Economic Journal 14 (2): 1-40. Wolak, F. A., and R H. Patrick (2001). "The Impact of Market Rules and Market Structure on the Price Determination Process in the England and Wales Electricity Market." NBER Working Papers, 8248. Wolak, F.A. (2003a). “Measuring Unilateral Market Power in Wholesale Electricity Markets: the California Market, 1988-2000.” American Economic Review 93 (2): 425430. Wolak, F.A. (2003b). “Diagnosing the California electricity crisis.” The Electricity Journal, Aug/Sept: 11-37. Wolak, F.A. (2010), “An Experimental Comparison of Critical Peak and Hourly Pricing: The PowerCentsDC Program”, Stanford University, Stanford, CA 94305-6072, 13
http://www.stanford.edu/~wolak
Fig 4.1 Hourly profile of electricity demand in Italy – annual average
HOURLY DEMAND MW/H: JUNE
4
6
x 10
2004 2005 2006 2007 2008 2009 2010 2011
5.5
5
4.5
MW/h 4
3.5
3
2.5 0
5
10
h
15
20
25
Quantity demanded on the third Wednesday of June for the years 2004-2011.
4
5.5
x 10
HOURLY DEMAND MW/H: DECEMBER
Quantity demanded on the third Wedsneday of June for the years 2004-2010. 5
4.5
2004 2005 2006 2007 2008 2009 2010
MW/h 4
3.5
14 3
Table 4.1 Main operators in the Italian electricity market – demand side Main Buyers Jun-08
Dec-08 Share BUYER %
Jun-09 Share BUYER %
Share %
GRTN
31,9
GRTN
28,6
30,2
SINGLE BUYER (AU S.P.A.)
21,4
SINGLE BUYER (AU S.P.A.)
27,1
ENEL TRADE S.P.A.
16,6
ENEL PRODUZIONE 17,4 S.P.A.
SINGLE BUYER (AU S.P.A.)
EDISON S.P.A. A2A S.R.L.
EDISON S.P.A. A2A S.R.L.
BUYER
EDISON S.P.A. A2A S.R.L.
TRADING TRADING
4,5 2,9
TRADING TRADING
4,6 2,5
Dec-09 Share BUYER % Bilateral GRTN 33,6 Agreement ENEL PRODUZIONE SINGLE BUYER 18,7 S.P.A. (AU S.P.A.)
TRADING TRADING
17,4 5,0 3,4
23,9
ENEL PRODUZIONE 17,1 S.P.A. EDISON S.P.A. A2A S.R.L.
TRADING TRADING
5,7 3,2
ENEL PRODUZIONE 2,1 S.P.A.
SORGENIA S.P.A.
2,0
SORGENIA S.P.A.
2,6
SORGENIA S.P.A.
2,6
Others
Others
17,8
Others
19,3
Others
17,3
20,7
Jun-10
Dec-10
Jun-11
BUYER
Share % BUYER
Share % BUYER
Share %
Bilateral Agreement
40,8
Bilateral Agreement
39,2
Bilateral Agreement
45,9
ACQUIRENTE UNICO S.P.A.
14,1
ACQUIRENTE UNICO S.P.A.
16,1
ACQUIRENTE UNICO S.P.A.
14,0
ENEL PRODUZIONE S.P.A.
12,9
ENEL PRODUZIONE S.P.A.
11,9
ENEL PRODUZIONE S.P.A.
7,7
EDISON TRADING S.P.A.
7,3
EDISON TRADING S.P.A.
6,5
EDISON TRADING S.P.A.
5,7
A2A TRADING S.R.L.
3,6
A2A TRADING S.R.L.
3,7
A2A TRADING S.R.L.
3,2
SORGENIA S.P.A.
3,1
SORGENIA S.P.A.
2,3
ENERGETIC SOURCE SPA
2,7
Others
18,2
Others
20,2
Others
20,8
15
Fig 4.2: Demand Market Shares of the main buyers (Jun08)
16
17
Table 4.2 Electricity prices in Italy. Monthly average price. Monthly average quantity demanded. Monthly liquidity ratio – euro /MWh 2004
2005
Period
National Average PUN (€/MWh)
Price
Total Quantity (MWh)
Liquidity Ratio (%)
National Average Price PUN (€/MWh) Total Quantity Liquidity (MWh) Ratio (%) Averag Min Max e
Average Min
Max
Jan
-
-
-
-
-
61,03
10,42
150,07
26.759.805
66,9
Feb
-
-
-
-
-
59,77
21,8
120,59
25.916.930
64,8
Mar
-
-
-
-
-
56,58
21,43
120,53
27.983.642
64,7
Apr
48,19
2,12
85,42
24.293.230
30,5
48,79
21,4
105
25.659.376
61,4
May
43,98
1,1
84,43
24.677.785
27,5
47,32
21,24
90,03
26.148.988
61
Jun
61,81
12,84 163,49
25.215.390
28,1
54,81
21,4
129,83
26.682.953
62,2
Jul
61,38
10,6
189,19
28.297.095
30,1
65,25
21,2
170,61
28.621.585
62,3
Aug
49,29
3,48
135,67
23.444.880
31,6
56,23
21,27
111,11
24.094.089
65,9
Sep
53,87
3,05
164,56
26.420.262
27,5
63,26
21,26
108,59
27.012.746
60,3
Oct
50,01
24,29 101,65
26.507.786
31,2
61,55
20,31
106,07
27.370.290
58,3
Nov
46,84
23,62 175,44
26.229.714
26,1
62,8
20,3
156,85
27.551.802
60,9
Dec
49,22
8,61
26.485.841
29,1
65,59
22
125,77
29.382.645
65,1
178,17
2007
National Average Price PUN (€/MWh) Total Quantity Liquidity (MWh) Ratio (%) Average Min Max
National Average Price PUN (€/MWh) Total Quantity Liquidity (MWh) Ratio (%) Averag Min Max e
Jan
72,3
22,61
170,74
29.294.381
64,6
76,34 21,63
190,22
28.558.473
65,6
Feb Mar Apr May Jun Jul Aug Sep Oct Nov
81,98 78,99 67,41 67,41 72,27 84,49 74,39 76,62 71,31 74,01
19,57 38,92 29,45 26,95 25,51 16,22 31,09 29,15 15,06 28
153,64 199,27 148,04 150,42 181,25 378,47 165,2 170,79 154,73 188,81
27.454.883 29.316.890 25.047.752 26.439.985 27.165.325 29.647.624 24.036.593 27.370.317 27.821.259 27.937.988
63,2 59,6 58,5 56,3 57 59,1 62,2 57,2 56,7 59,3
69,77 61,47 55,5 63,03 67,16 83,83 63,01 69,84 69,86 90,82
173,96 143,18 111,93 137,09 185,29 209,8 178,06 179,35 191,95 242,42
26.045.519 27.642.958 25.300.179 27.241.732 27.641.265 30.778.170 24.765.879 26.840.918 28.257.218 28.308.879
64,5 62,8 63,9 69,9 67,9 68,9 69,6 65,7 67,3 68,2
Dec
76,28
31,66
199,02
28.257.032
61,2
81,08 32,5
214,65
28.568.020
70
Period
2006
22 22 22 22,04 22,17 21,67 21,52 22 21,44 28,09
18
Period
2008
2009
National Average Price Total PUN (€/MWh) Quantity (MWh) Averag Min Max e
National Average Price PUN (€/MWh) Total Quantity Liquidity (MWh) Ratio (%) Average Min Max
Liquidity Ratio (%)
Jan
86,24
29
196,09
29.382.901 69,5
83,45
23,56
168,07
27.071.149
68,6
Feb
81,49
29
191,11
28.037.468 67,3
76,95
23,6
125,46
24.984.931
66,3
Mar
74,54
29
161,52
28.279.622 67,4
69,1
23
163,07
26.685.863
66,1
Apr
80,62
29
175
26.608.515 65
58,36
11,7
158,45
24.339.903
67,8
May
80,09
26,05
161,55
27.089.030 66
58,51
10,56
135,19
24.998.492
67
Jun
83,49
23,48
185,75
28.536.349 68,2
51,82
10
136,73
25.103.967
67,1
Jul
97,32
23,9
211,99
31.113.127 70,7
60,5
16,25
149,66
28.670.111
68,5
Aug
90,95
29,56
178,82
25.340.986 72,6
71,07
20,19
169,38
23.600.054
70,9
Sep
97,23
29,52
163,19
28.374.248 71,2
66,49
16,24
172,25
26.688.855
68
Oct
99,07
29,53
173,58
28.892.110 69,2
57,63
14
139,36
28.117.122
67,5
Nov
87,65
24,42
160,15
27.834.750 70
53,93
13,49
136,01
26.444.819
68,1
Dec
84,87
21,54
172,09
27.472.189 71,4
57,39
10,35
147,87
26.719.900
69,8
2010
2011
Liquidity Ratio (%)
Average Min
Price Total Quantity (MWh) Max
Jan
63,45
10
174,62
27.447.219
Feb
62,55
20,54
123,3
Mar
62,82
18,07
Apr
61,31
22,82
May
59,36
Jun
Period
National Average PUN (€/MWh)
National Average PUN (€/MWh)
Price Total Quantity Liquidity (MWh) Ratio (%)
Average
Min
Max
64,1
65
10
91,72
27.189.124
59,5
26.056.441
65,8
66,29
27
110,4
25.488.246
56,7
118,57
27.657.283
64,6
68,18
36,18
142,96
27.023.464
55,9
117,54
25.162.019
63,1
65,18
22,94
118,07
24.059.769
57,9
15,88
100,43
25.811.832
61
71,28
39,98
98,01
25.240.480
58,7
60,2
12,87
120,94
25.793.643
62,6
68,41
37,49
102,83
25.537.893
59,9
Jul
70,9
23,55
157,52
28.827.844
62
69,74
16,14
142,67
28.448.304
57,6
Aug
69,91
23,55
145,01
24.301.309
60,9
74,51
31,06
132,99
24.095.662
56
Sep
66,55
24,95
140,61
26.073.948
60,9
81,31
40,01
131,71
26.149.600
57,6
Oct
65,78
25,29
138,88
27.028.151
60,5
78,61
23,53
151,09
26.183.973
58,2
Nov
61,38
22,67
162,98
26.657.773
62,5
78,47
28
160,62
25.928.261
58,1
Dec
64,88
22,69
118,21
27.744.102
63,1
79,37
28
164,8
26.149.101
58,4
19
Table 4.3 Electricity prices in Italy – monthly average - euro /MWh
Period
Monthly Average National Single Price
2004
2005
2006
2007
2008
2009
2010
2011
Jan
-
61,03
72,30
76,34
86,24
83,45
63,45
65,00
Feb
-
59,77
81,98
69,77
81,49
76,95
62,55
66,29
Mar
-
56,58
78,99
61,47
74,54
69,10
62,82
68,18
Apr
48,19
48,79
67,41
55,50
80,62
58,36
61,31
65,18
May
43,98
47,32
67,41
63,03
80,09
58,51
59,36
71,28
Jun
61,81
54,81
72,27
67,16
83,49
51,82
60,20
68,41
Jul
61,38
65,25
84,49
83,83
97,32
60,50
70,90
69,74
Aug
49,29
56,24
74,39
63,01
90,95
71,07
69,91
74,51
Sep
53,87
63,26
76,62
69,84
97,23
66,49
66,55
81,31
Oct
50,01
61,55
71,31
69,86
99,07
57,63
65,78
78,61
Nov
46,84
62,80
74,01
90,82
87,65
53,94
61,38
78,47
Dec
49,22
65,59
76,28
81,08
84,87
57,39
64,88
79,37
Average Price (PUN)
51,62
58,58
74,79
70,97
86,96
63,77
64,09
72,20
20
Table 4.4 Number of zones in Italian electricity market – relative frequency Frequency of different zone aggregation 2004
2005
2006
2007
2008
2009
2010
2011
National Single Market
4,85%
23%
19%
22,55%
19,35%
15,18%
17,7%
15,5%
Two Zones
27,85%
47,75%
40,7%
41,75%
44,3%
35,6%
37,8%
45,9%
Three Zones
46,3%
25,80%
29;9%
29,5%
29;5%
37,5%
32,45%
31,61%
Four Zones
19,4%
3,35%
9,25%
6%
6,2%
11,26%
11,35%
6,56%
Five or More Zones
1,6%
0,1%
1,15%
0,2%
0,65%
0,46%
0,7%
0,43%
Total
100%
100%
100%
100%
100%
100%
100%
100%
Table 4.5 Zonal Electricity prices in Italy – annual average - euro /MWh Annual Average Zone Price Zone Northern Italy Central-Northern Italy
2004
2005
2006
2007
2008
2009
2010
2011
54,2
57,71
73,67
68,46
82,91
60,88
61,95
70,15
53,06
58,62
75,01
72,78
84,98
62,31
62,43
71,13
Central-Southern Italy Southern Italy
54,2
59,03
75,02
73,03
87,59
62,45
62,56
70,82
54,2
59,03
75,01
73,03
87,36
59,52
58,97
68,99
Sicilia
55,24
62,74
78,95
79,50
119,51
88,05
89,77
93,01
Sardegna
59,88
60,36
80,57
74,99
91,74
82,06
73,51
79,86
Calabria
56,42
59,82
75,70
73,22
87,95
-
-
-
Annual Average Price
55,31
59,61
76,27
73,57
91,72
69,21
68,19
75,66
21
Table 4.6 Aggregate price elasticity of electricity in Italian market – estimates from GAI function, eq 5
HOURLY DEMAND ELASTICITY
2011
2010
MOUNTH
MAX
MIN
AVERAGE
JANUARY
-0.2640
-0.0117
-0.0733
FEBRUARY
-0.3270
-0.0076
-0.0854
MARCH
-0.281
-0.0075
-0.0320
APRIL
-0.212
-0.0327
-0.0825
MAY
-0.256
-0.0077
-0.0980
JUNE JULY AUGUST SEPTEMBER
-0.227
-0.0082
-0.0881
-0.161
-0.00025
-0.0397
-0.153
-0.0118
-0.0612
-0.232
-0.0117
-0.0790
OCTOBER
-0.245
-0.0085
-0.0747
NOVEMBER ECEMBEDR JANUARY FEBRUARY MARCH APRIL MAY JUNE JULY AUGUST SEPTEMBER
-0.30
-0.018
-0.0671
-0.135
-0.0058
-0.0290
-0.132
-0.00028
-0.0206
-0.209
-0.0003
-0.0806
-0.218
-0.0039
-0.0758
-0.2430
-0.0003
-0.0399
-0.0629
-0.0064
-0.0252
-0.0406
-0.00019
-0.0208
-0.0327
-0.00286
-0.0190
-0.0337
-0.00535
-0.0179
-0.0462
-0.00439
-0.0130
22
Table 4.7 Demand price elasticity of electricity in Italian market – by day of the week
AVERAGE ELASTICITY BY DAY OF THE WEEK
2011
2010
MON I QAURTER 2010 II QUARTER 2010 III QUARTER 2010 IV QUARRTER 2010 I QUARTER 2011 II QUARTER 2011 III QUARTER 2011
TUE
WEN
THU
FRI
SAT
SUN
-0.0477
-0.0808
-0.0945
-0.0477
-0.0789
-0.0660
-0.0442
-0.0667
-0.0976
-0.1142
-0.1032
-0.1066
-0.0836
-0.0555
-0.0602
-0.0678
-0.0565
-0.0674
-0.0626
-0.0533
-0.0375
-0.1022
-0.1269
-0.1047
-0.1231
-0.1304
-0.1207
-0.0877
-0.0409
-0.0549
-0.0710
-0.0689
-0.0636
-0.0628
-0.0465
-0.0194
-0.0329
-0.0333
-0.0342
-0.0365
-0.0237
-0.0191
-0.0163
-0.0158
-0.0158
-0.0182
-0.0185
-0.0200
-0.0120
23
Table 4.8 Demand price elasticity of electricity in Italian market – peak and off peak hours
AVERAGE ELASTICITY: PEAK – OFF PEAK
PEAK (1112-1314) OFF_PEAK (23-241-2)
I QUARTER 2010
II QUARTER 2010
III QUARTER 2010
IV QUARTER 2010
I QUARTER 2011
II QUARTER 2011
III QUARTER 2011
-0.09065
-0.08383
-0.06578
-0.05671
-0.06867
-0.02873
-0.01386
-0.05018
-0.09439
-0.05634
-0.05457
-0.05067
-0.02859
-0.01698
-0.09733
-0.09864
-0.07153
-0.05801
-0.06693
-0.03153
-0.01411
-0.05644
-0.10826
-0.06118
-0.05783
-0.05411
-0.02917
-0.01756
PEAK NOFESTIVI OFF - PEAK NO FESTIVI
24
Table 4.9 Demand price elasticity of electricity in Italian market – average value for each marginal operator determining the equilibrium price
MARGINAL BUYERS – AVERAGE ELASTICITY (2010) ALPIQ ENERGIA ITALIA S.p.A. ACEAELECTRABEL TRADING S.P.A. ALPIQ SA ALPIQ SUISSE SA AXPO AG AZIENDA ELETTRICA TICINESE BKW FMB ENERGIE AG BURGO ENERGIA S.R.L. COMPAGNIE NATIONALE DU RHONE (CNR) DANSKE COMMODITIES A.S DEUTSCHE BANK AG LONDON BRANCH E AND T ENERGIE HANDELSGESELLSHAFT mbH E.ON ENERGY TRADING S.P.A. EDF TRADING LIMITED EGL AG EGL-ITALIA S.P.A. ELECTRADE SPA ELECTRADE SRL ELEKTRIZITATS-GESELLSCHAFT LAUFENBURG AG ENERGI DANMARK A/S ENI SPA EXEN S.R.O EZPADA S.R.O. GEN-I TRGOVANJE IN PRODAJA ELEKTRICNE ENERGIJE D.O.O. GEOENERGIE S.P.A. HSE D.O.O. IBERDROLA GENERACION S.A.U. IMC ENERGY TRADING B.V. IRIDE MERCATO SPA J.ARON AND COMPANY GENERAL PARTNERSHIP LUCAS ENGINE S.P.A. MERRILL LYNCH COMMODITIES (EUROPE) LIMITED NECO S.A. NORDJYSK ELHANDEL A/S STATKRAFT MARKETS GMBH TEI ENERGY S.p.A. TIRRENO POWER S.P.A. VERBUND -AUSTRIAN POWER TRADING ENERGA HELLAS S.A.
-0.0636
-0.04415 -0.05915 -0.06648 -0.07135 -0.05998 -0.06078 -0.035952 -0.056658 -0.07019 -0.07851 -0.0335 -0.0373507 -0.07710 -0.03323 -0.07918 -0.05066 -0.072159 -0.03890 -0.06865 -0.0254 -0.07515 -0.08210 -0.0492
-0.01843 -0.16 -0.06113 -0.1345 -0.02674 -0.03505 -0.07353 -0.08315 -0.70328 -0.00667 -0.07136 -0.04965 -0.03015 -0.09278 25
MARGINAL BUYERS – AVERAGE ELASTICITY (2010) AGSM ENERGIA S.P.A. ALPIQ SA ALPIQ SUISSE SA AXPO AG AZIENDA ELETTRICA TICINESE AZIENDA ENERGETICA TRADING S.R.L.
-0.02234 -0.04124 -0.03843 -0.03915 -0.03815 -0.01145
BKW FMB ENERGIE AG BKW ITALIA SPA BLUE AEGEAN ENERGY S.A. C.U.RA. CONSORZIO UTILITIES RAVENNA S.C.R.L. CENTRALSCHWEIZERISCHE KRAFTWERKE AG (CKW AG) COMPAGNIE NATIONALE DU RHONE (CNR) DANSKE COMMODITIES A.S DEUTSCHE BANK AG LONDON BRANCH E.ON Energy Trading SE Sede Secondaria Italiana EDELWEISS ENERGIA S.P.A. EDF TRADING LIMITED EGL AG EGL-ITALIA S.P.A. ELECTRADE SPA ELPEDISON TRADING S.A. ENEL PRODUZIONE S.P.A. ENERGA POWER TRADING SA ENERGI DANMARK A/S ENERGY FINANCING TEAM AG ESPERIA S.P.A EXEN S.R.O GDF SUEZ ENERGY MANAGEMENT S.p.a. IBERDROLA GENERACION S.A.U. IREN MERCATO SPA LUCAS ENGINE S.P.A.
-0.03403 -0.01337 -0.0294 -0.01865 -0.02826 -0.03061 -0.06153 -0.02472 -0.02945 -0.02146 -0.06713 -0.0421 -0.03691 -0.04053 -0.01877
NECO S.A. ONDA SRL PUBLIC POWER CORPORATION S.A. SORGENIA S.P.A.
-0.04331 -.04237 -0.01855
STATKRAFT MARKETS GMBH TEI ENERGY S.p.A. TIRRENO POWER S.P.A.
-0.04502 -0.1842
-0.0147
-0.02819 -0.0281 -0.0304
-0.01850 -0.02325 -0.01846 -0.0819 -0.02514 -0.03559
-0.19
-0.02061
Table 26
4.10 Demand price elasticity of electricity in Italian market – average value for each market zonal division
AVERAGE ELASTICITY: ZONE DIVISION I
II
III
IV
I
II
III
QUARTER QUARTER QUARTER QUARTER QUARTER QUARTER QUARTER (2010) (2010) (2010) (2010) (2011) (2011) (2011) 1 ZONE
-0.0457
-0.0874
-0.05368
-0.0534
-0.04543
-0.02844
-0.01713
N. OBS
(356)
(440)
(399)
(355)
(414)
(228)
(217)
2 ZONES
-0.0637
-0.0865
-0.06008
-0.06351
-0.05474
-0.02603
-0.01706
N. OBS
(777)
(858)
(825)
(848)
(1116)
(941)
(820)
3 ZONES
-0.08027
-0.0942
-0.06147
-0.05637
-0.07295
-0.03084
-0.01622
N. OBS
(695)
(655)
(783)
(710)
(564)
(804)
(924)
4 ZONES
-0.09419
-0.09010
-0.06371
-0.04250
-0.07513
-0.03238
-0.01666
N. OBS
(318)
(206)
(199)
(274)
(65)
(195)
(959)
-0.0449
-0.11249
-0.0784
-0.04865
-
-0.02120
-0.01986
N. OBS
(13)
(25)
(2)
(21)
-
(16)
(9)
TOT.
2159
2184
2208
2208
2159
2184
2208
5 OR MORE
ZONES
27
Table 4.11 System marginal price distribution – average values by deciles
SMP: SUMMARY STATISTICS 2010
2011
N. OBS
MEAN
STD. DEV
MIN
MAX
N. OBS
MEAN
STD. DEV
MIN
MAX
876 876 876 876 876 876 876 876 876 876
32.97 45.81 52.24 58.42 63.04 66.55 69.69 73.78 80.01 98.70
5.92 2.30 1.8 1.62 1.15 0.88 0.97 1.33 2.43 16.04
10 40.94 49.42 55.39 61.1 65.05 68.03 71.6 76.24 84.94
40.93 49.40 53.38 61 65 68.02 71.5 76.24 84.93 174.62
656 655 655 655 655 655 655 655 655 655
44.40 55.31 61.59 65.66 68.44 71.05 74.20 78.03 83.75 97.66
6.57 2.20 1.40 1.01 0.7 0.83 0.93 1.36 2.01 9.7
10 51.24 58.98 63.79 67.19 69.70 72.61 75.93 80.64 87.70
51.21 58.97 63.79 67.18 69.69 72.60 75.92 80.63 87.57 142.96
28
Table 4.12 Demand price elasticity of electricity in Italian market – average value for each decile of the equilibrium system marginal price (SMP) distribution
ELASTICITY BY SMP PERCENTILES 2010 MEAN
MAX
2011 MIN
MEAN
MAX
MIN
-0.00755
-0.03007
1
-0.0562186 -0.264
-0.199
-0.000277
2
-0.0596927 -0.242 -0.000248 -0.03318005 -0.239
-0.000237
3
-0.0598467 -0.277 -0.000308
-0.0327781
-0.241
-0.000235
4
-0.0692816 -0.294 -0.000302
-0.0475505
-0.226
-0.00225
5
-0.0711693 -0.239 -0.000309
-0.0415078
-0.243
-0.000238
6
-0.0808743 -0.327
-0.00029
-0.0339698
-0.201
-0.000204
7
-0.0813875
-0.00585
-0.0343407
-0.209
-0.000193
8
-0.0786965 -0.266 -0.000322
-0.0316278
-0.217
-0.000198
9
-0.0707198 -0.287 -0.000313
-0.0318237
-0.183
-0.000189
-0.0632563 -0.253
-0.0283154
-0.218
-0.00116
10
-0.3
-0.00596
29
0
5
Density
10
15
HOURLY ELASTICITY DISTRIBUTION (JAN-DEC 2010)
-.327
-.000248 ELASTICITA
0
10
Density 20
30
40
HOURLY ELASTICITY DISTRIBUTION (JAN – SEP 201)
-.243
-.000189 ELASTICITA
30
