Linkoping Studies in Science and Technology Thesis No.746

c, MARIU till 4 R I.T., Optimized Systems and New Customer Services

S' > The Deregulated Electricity Market and the Ronneb: ‘EllVED APR 3 01999 Ulfika BergstroifltS^I /Six -aL W INSTITUTE OF TECHNOLOGY LINKOPING $ UNIVERSITET

LiU-TEK-^IC-1999:04

Division of Energy Systems Department of Mechanical Engineering Linkopings Universitet, S-581 83 Linkoping, / Linkoping 1999 Linkoping Studies in Science and Technology Thesis No. 746

I.T., Optimized Energy Systems and New Customer Services

The Deregulated Electricity Market and the Case % Ulrika Bergstrom

Akademisk avhandling

som for avlaggande av teknologie licentiatexamen vid Linkopings tekniska hogskola kommer att offentligt forsvaras i H&llfasthetslara’s seminarierum, hus A, ing&ng 15 Linkopings universitet, mindagen den 19 april 1999 kl 13.00. Opponent ar doktor Maria Andersson, SAAB, Linkoping. Avhandlingen finns tillganglig vid amnesomr&det Energisystem, IKP, hus A, Linkopings Universitet

Abstract This thesis concerns the utilization of information technology (IT) to obtain optimized energy systems and the increasing dependence on IT applications within the power industry in general, and especiallywithin electricitysales. i Based on energy systemoptimizations, industrial simulations, interviews and literature "i surveys, the thesis concludes that information technology is a condition for optimized energy systems. Through large-scale load control, enabled by IT, the energy system cost for supplying a local energy system can be reduced considerably.

Diurnal energy systemoptimizations further illustrate the increased utilization of load control, accentuated, when customers are exposed to the variations of the spot market price. Thereby, the increased need for load management on a deregulated market, where real time pricing is applied, is mirrored.

However, given the boundary conditions of the deregulated electricity market, optimization of energy systemsis no condition for competitiveness within electricity sales. This study also points out that, apart from industrial customers, other market actors, like electricitysales companies and local distributors, have few incentives to introduce load control. Distributors mainly lose money on power reducing measures, except for when there are distribution limitations. Neither do electricity sales companies make short-term earnings from large-scale . : i load control. The combination with small economic savings for residential customers make the financing of large-scale load control problematic.

Regarding electricitysales companies, increased IT utilization is observed to enable marketing and service offerings on the Internet, and IT-based value-added services (VAS). For pure communication services, the sales companies utilize IT to offer telephone services (analogue technique) and Internet access, the latter is sometimes performed on the electricity grid. Extended IT-systems for customer administration, customer service, and market analyses are consequences of the competition and of aggressive corporate energy expansion. :l Due to the introduction of other IT-based services, load control is likely to become utilized as well, since this information flow might piggy-back on the costs of other more attractive services for the customers. Thereby, as a side effect, the deregulation may also lead to more i optimized energy systems.

Division of Energy Systems Department of Mechanical Engineering Linkopings Universitet, S-581 83 Linkoping, Sweden ISBN 91-973567, ISSN 0280-7971

.}< ‘i'V Linkoping Studies in Science and Technology Thesis No.746

I.T., Optimized Energy Systems and New Customer Services

The Deregulated Electricity Market and the Ronneby Case

Ulrika Bergstrom

/ST* % v INSTITUTE OF TECHNOLOGY LINKOPINGS UNIVERSITET

LiU-TEK-LIC-1999:04

Division of Energy Systems Department of Mechanical Engineering Linkopings Universitet, S-581 83 Linkoping, Sweden

ISBN 91-973567-1-9 ISSN 0280-7971 DISCLAIMER

Portions of this document may be illegible in electronic image products. Images are produced from the best available original document. ABSTRACT

This thesis concerns the utilization of information technology (IT) to obtain optimized energy systems and the increasing dependence on IT applications within the power industry in general, and especially within electricity sales.

Based on energy system optimizations, industrial simulations, interviews and literature surveys, the thesis concludes that information technology is a condition for optimized energy systems. Through large-scale load control, enabled by IT, the energy system cost for supplying a local energy systemcan be reduced considerably.

Diurnal energy system optimizations further illustrate the increased utilization of load control, accentuated, when customers are exposed to the variations of the spot market price. Thereby, the increased need for load management on a deregulated market, where real time pricing is applied, is mirrored.

However, given the boundary conditions of the deregulated electricity market, optimization of energy systems is no condition for competitiveness within electricity sales. This study also points out that, apart from industrial customers, other market actors, like electricity sales companies and local distributors, have few incentives to introduce load control. Distributors mainly lose money on power reducing measures, except for when there are distribution limitations. Neither do electricity sales companies make short-term earnings from large-scale load control. The combination with small economic savings for residential customers make the financing of large- scale load control problematic.

Regarding electricity sales companies, increased IT utilization is observed to enable marketing and service offerings on the Internet, and IT-based value-added services (VAS). For pure communication services, the sales companies utilize IT to offer telephone services (analogue technique) and Internet access, the latter is sometimes performed on the electricity grid. Extended IT-systems for customer administration, customer service, and market analyses are consequences of the competition and of aggressive corporate energy expansion.

Due to the introduction of other IT-based services, load control is likely to become utilized as well, since this information flow might piggy-back on the costs of other more attractive services for the customers. Thereby, as a side effect, the deregulation may also lead to more optimized energy systems.

i The thesis contains two main presentations.

The first focuses on the main tendencies regarding information technology utilization within the different areas of power industry, concentrating on the impact of deregulation. In this part, there is an introduction, containing background and hypothesis, a study of consequences from deregulation on different actors, a survey of IT utilization in the power industry and a description of the performed work and the methodology.

In the second part, a case study of the Ronneby energy system is outlined, which referring to the enclosed papers, shows the necessity of information technology in obtaining an optimized energy system. Finally, comments on the enclosed papers appear and conclusions are drawn.

The following papers are included in thethesis:

(I) Bergstrom, U., Simulation of a local energy system with focus on cost efficient DSM measures on a deregulated electricity market, In the proceedings of the international conference on Distribution Automation and Demand Side Management (DA/DSM), October, Vienna, Austria, pp 631-644, 1996.

(II) Bergstrom, U., Impact of deregulation - DSM in industries, In the proceedings of the international conference on Distribution Automation and Demand Side Management (DA/DSM), October, , UK, CD ROM, 1998.

(III) Bergstrom, U., Traditional load management versus real time load control, Paper for the European (EU) RISI (Regional Information Society Initiative) conference, November, Graz, Austria, 1998.

II ACKNOWLEDGEMENTS

Many creative discussions with several inspiring persons have contributed to the completion of this thesis.

To start with, I want to thank my supervisor Prof. Bjorn Karlsson, for his support and advice during this time. I also gratefully acknowledge the financial support from Enersearch and Sydkraft that has enabled the writing of this thesis. Especially, I would like to thank Hans Ottosson, Enersearch, for his never-ceasing enthusiasm regarding IT and energy, and for bringing inspiration into the project at times when my creativity was at low.

Further, I would like to thank my employer, SYCON, for giving me the opportunity to enter more deeply into the questions of energy system optimization.

Special thanks go to Tekn. Dr. Dag Henning, for all his valuable advice and especially for his patience with all my never-ending questions, put to him at the most inconvenient times, weekdays and weekends.

I would also like to express my gratitude to Hans Olausson, President of Ronneby Energi AB, who has introduced me to the business of REAB, the local electricity distribution utility, focused on in part of this thesis. Warm thanks, also to Lennart Nilsson, Malmo Varme, for many pleasant conversations, and for sharing some of his experience of district heating operation.

Many thanks go to my colleagues at SYCON, especially to Lennart Larsson, Per-Axel Nilsson and Bertil Strandh, who always took time off to discuss certain issues. At Sydkraft, I especially would like to thank Paul Lundgren and Mats Jonsson, Sydkraft Elforsaljning, Bengt Ahlman, Sydkraft Utveckling, Kenneth Johansson, Sydkraft Eldistribution, and Torbjom Forsberg, Bengt-Thore Sondh and Bengt Persson, Sydkraft Produktion, for valuable viewpoints regarding the utilisation of information technology within their respective areas.

I also wish to thank my parents and parents-in-law, for all their support and for taking care of our entire family, Fredrik, Leif and me, at times when the work load was too heavy.

Finally, last but not least, my love and my deepest thanks to Leif and Fredrik for putting up with these intensive years of chaos.

Lund, 9 February 1999

Ulrika Bergstrom

m 1.3 Purpose The purpose of this thesis is to study information technology as a condition for optimized energy systems, the consequences on load control due to the deregulation of the electricity market and the development of IT-related value added services.

2. CONSEQUENCES OF THE DEREGULATED ELECTRICITY MARKET The aim of the deregulation of the electricity market was mainly to reduce all customers ’ electricity prices. Through competition, the power industry would be forced to increase production efficiency and reduce margins of the electricity sales. Therefore the V.th of January 1996 the Swedish electricity legislation, originally from 1902, was reformed 2, the Nordic Power Exchange Nord Pool 3 4was inaugurated and thereby the electricity market finally opened for competition.

2.1 Legal aspects Major changes were introduced upon the power industry. Below follows a summary of the most important stipulations relevant for the content of this thesis.

• The electricity sales and the electricity production were separated from the transmission and distribution of electricity.

• The distribution stayed in a monopoly, while the sales and production went competitive.

• The distributor may only own electricity production capacityas large as the losses in his own distribution grid.

• All distribution must be open for all actors as long as they pay a point tariff".

• The distribution companies are obliged to transmit power, sold by any trading company, to their customers. They are also responsible for electricity measurements and reporting within their concession area.

• All customers have the right to change electricity supplier. To do so, hourly measurements are required. This requirement will, however, be taken away in November 1999.

• A certain electricity delivery concession guarantees electricity deliveries of at least five years for customers satisfied with their ordinary supplier.

In the beginning, the Governmental Authority'of Energy controlled and supervised the monopoly of the distribution to secure that customers got equal conditions for their

2 The old legislation was complemented with the law of electricity trading. Later, in January 1998, these two laws were replaced as the new electricity legislation (1997:857) came into force. 3 Before 1996, the power exchange only managed Norwegian electricity trade. 4 This means that when the fee is paid in accordance with the stamp principle for the use of the distribution lines, the actor/customer gets access to the whole distribution grid. 2 services. Now, the Governmental Authority of Electricity Distribution has taken over that role. The Swedish Competition Authority regulates the competitiveness of the actors of the electricity sales market. Svenska Kraftnat has, among other responsibilities, the overall liability for keeping the short-term balance of the national power system.

2.2 Structural Changes On themonopoly electricity market, within a region, one single actor, like Vattenfall or Sydkraft, could dispose all assets required for the electricity supply, namely the production units, the regional distribution grid and sometimes also the local distribution. Today, several actors might operate within the same concession area and, therefore, higher demands are put on, for example, metering, deduction and supply forecasts (see 3.1.1). Another consequence is increased difficulty in optimization of local or regional energy systems by means of only focusing on one actor. Earlier the business relations of an energy system could briefly be illustrated by the PDC concept, see figure 2.1. The producer ’’pushed” his delivery out to the local energy utility, whichsupplied the final end-consumers.

Producer

Figure 2.1 Business relations within a region before the deregulation5

Apart from the possibility of lowered electricity prices, the deregulation has brought increased complexity upon the customers. For residential customers at least two business relations have to be handled: one with the electricity sales company and one with the local, often municipal, distributor. Industries and corporations sometimes have more than three contacts: one or several with local distributors (due to location), one or more with electricity sales companies, and sometimes also contacts with electricity brokers or directly with the power exchange. To divide their dependence and risks amongst different market players, some of these customers also seek for shared delivery contracts. Looked upon as customers, the same conditions are applicable to municipal electricity sales companies. In figure 2.2 potential new business relations of thederegulated market are illustrated. * 3

5 The regional distributor is not included in the picture. 3 POWER EXCHANGE

ESC 1 BROKER l...n

LOCAL ESC

CUSTOMER LOCAL DISTRIBUTOR

REGIONAL DISTRIBUTOR ____ Monopoly business relations ____ Potential business relations NATIONAL of the free market TRANSMITTOR

Figure 2.2 Examples of potential business relations within the electricity market

Besides the increased complexity also the direction of the trade initiatives among the market actors has partly changed. At the monopoly market, customers had few possibilities to influence their situation or the behaviour of the supplier. But, since the market went competitive, suddenly, the customers have taken the initiative to gain new advantages, in terms of reduces prices or increased services, by playing the electricity suppliers off against each other. Thereby, the deregulation has partly turned the electricity market from a push to a pull market.

With the introduction of the free market the three activities, electricity production, distribution and sales have obtained totally different driving forces and are run according to totally different parameters of success. Power generation and electricity sales are both activities exposed to competition. Therefore, they concentrate on increasing their respective margins of sales versus the marginal production cost or the purchase costs 6. Provided, that the businesses are separated, apart from the electricity sales companies, the maximum price possible that the producer can obtain for the supply is official and settled by the power exchange. Thus, several measures are performed to increase efficiency to reduce the short range marginal production cost (SRMC) and thereby also secure the sale. For example, in Sweden, several cost saving programs have been performed in terms of centralization and general rationalization. But, when such efficiency improvements are not enough to reduce the SRMC so that a continuous market for the deliveries can be obtained, the production units are decommissioned. This

6 In opposite, monopoly businesses are known to, to a greater extent than competitive companies, also include non profit- related goals in their objectives [42]. has already happened to some peak production units in Sweden, like those in Stenungsund and .

Regarding electricity sales companies, their focus is on innovative purchase strategies, in order to keep down the electricity purchase costs from the power exchange, from the brokers or from the production units within the own combine group. This way they try to keep or increase the margins from their sales. Further, some of the companies provide value added services (VAS) in an effort to defend the price level. Therefore, the electricity sales companies have little interest in subsidizing combine group internal, but costly production units.

As part of the VAS strategy, the sales companies also strive to obtain good customer relations and to become customer representatives. For example, they try to integrate with the customers and to increase their sales by promise of lower energy costs. Their earning is the winning or retaining of the customer. A consequence of the strategy is the potential conflict it might lead to with the distributor.

In opposite, the monopoly distributor ’s objective is to distribute as much power as possible, without exceeding the limiting capacity of his own distribution grid or costly power-limitations contracted with the regional net operator. Since the Governmental Authority of Electricity Distribution supervise and regulate the distribution income, the distributors main measures to increase their yield consist of different types of efficiency improvements.

Regarding producers and sales companies, one of the main differences between monopoly and competitive market conditions, concerns their financial risk exposure. After the deregulation, a new risk situation has occurred for these companies. Earlier they had control both of the production capacity and of the sales which generated a very good cost control. The revenue risks were low, and they also had good control of the investment risk of delivery obligations. After the deregulation, the revenue risk increased considerably, since there on the one hand was an investment risk for market shares, and on the other hand, there was a higher liquidity risk. For example, if plant production suddenly dropped, the company would immediately be exposed to the regulating price at the Regulation Exchange market (see 2.2.1). Thesame would happen due to a sudden loss of market shares. Of course, this new risk situation also influences eventual investment plants for the building of new production units.

Another phenomenon that changes the traditional picture, is the emergence of entirely new market players, with no former connection to the power business. With the introduction of Nord Pool a new industry has formed, namely the electricity market makers, brokers only engaged in the daily buying and selling of power. Some of these have experience from other commodity exchanges. Banks, oil-conglomerates and insurance companies are traders at Nord Pool. Also OK, Shell and Staoil have started to sell electricity. So-called ” department stores of energy” have been created.

So far, the new actors mainly concentrate on large consumers. But a change, and a growth in the number of market participants, is to be expected, especially when the requirement for hourly metering for small customers is taken away.

5 At the same time as the number of new market actors increases, the number of traditional actors, like local distributors and sales companies, decreases. In February 1996, there were 227 electricity sales companies and 248 distributions companies. In the beginning of 1999 the figures were around 100 sales and 240 distribution companies. The figures for the sales companies have changed considerably and so has the business structure. In the beginning of 1998, many electricity sales companies were part of an electricity combine, together with a local distribution utility. But, in order to stay in business and to become competitive compared to the large producers and their sales organizations, an important trend is the joint-ownership of sales companies. In the beginning of 1998, there were twenty-four such companies, and the figure increases. The owners are mostly small or medium-small municipal energy utilities [73].

Another reason for the reduction of local distributors and sales companies is the offensive purchase and expansion strategies of larger energy corporations. To achieve advantages and gain market shares, companies like Sydkraft, Vattenfall and Stockholm Energi, have bought sales and/or distribution of local utilities. Examples of such advantages are the control of the households and the former company’s customer register, i. e. customer information. Corresponding development can be seen in the also deregulated electricity markets of California and the UK. In the UK and in the USA market players claim increased volume to be the key parameter for success in the competition. The American market has been estimated to room at the most ten dominating actors. The corresponding figure for the British market is said to be five [23].

Also interesting is that geographical conditions are of little importance. As Swedish groups started to act on the entire Nordic market, so did foreign companies begin to do business on the Swedish market. Major purchases and new alliances have changed the proprietor structure and electricity sales strategies in Sweden. The Finnish Imatran Voima, the German Preussen Electra, the French Electricite de France, the Norwegian Statkraft and Norsk Hydro, the Danish Elsam, the American Enron (natural gas), the British Eastern group (natural gas and electricity) and the Belgium Electrabel are only some examples of aggressive foreign companies that have established themselves in the Swedish market. Swedish electricityproducers have done the same, so far mainly in the Nordic countries. Sydkraft forecasts that the amount of multi-national actors will increase due to the globalization and to the hunt for large-scale operation advantages. At the same time, thenumber of national market operators will decrease.

Within Europe though, so far mainly the electricity markets of the Nordic Countries and the UK are deregulated. Thereby these markets also are of main interest for other market actors, of non-deregulated countries. The electricity markets within the European Union are said to become deregulated as well. According to the directive of joint regulations for the internal electricity market7, the member states shall realize the directive before February 1999 in national law. The aim is that the electricity market, step by step, will become open at the same time as public services are guaranteed. All countries are obliged to authorize customers to trade in a free market.

’(96/92/EG) 6 In contrast to the larger industries and companies, residential customers have not had the possibility of unconditionally changing their electricity supplier. Until November 1999, small customers will have to pay for the required electricity meter if they want to change supplier. Due to many households ’ relatively small electricity consumption, the meter cost takes away their eventual earnings from a supplier change.

Therefore, to avoid the regulatory restrictions limiting their freedom of action in electricity purchasing, customer alliances have been formed in the marketplace. Consumers have gathered their efforts to attain lower prices; house owners have gathered in Villaagareforeningen, municipalities have assembled in Kommentus, and, for example, the estate owner HSB and large industries have made combine group purchases.

2.2 Thepower exchange The Swedish-Norwegian power exchange market (Nord Pool) has a key role in the Nordic electricity power market. The exchange was established in January 1995, as a Norwegian exchange, and in January 1996, Sweden decided to participate in connection with the introduction of the deregulation. Nord Pool is owned by Svenska Kraftnat and Statnett SF with equal shares. It organizes both physical and financial electricity trade, as well as clearing services to their participants.

According to other deregulation acts in recent history, the number of market participants can be predicted to rise quickly. Such happened when the American trucking industry was deregulated: the number of brokers rose from 60 in 1980 to 4500 in 1985 [83]. Likewise, independent gas brokers filled the post-regulation void in the natural gas industry, so that the pipelines of gas carried by others doubled in eighteen months. Viewing the Nordic power exchange market, the tendency is similar. In 1997, Nord Pool had a turnover of more than six billion Norwegian crowns (NOK8 ). In 1998, this figure had increased to 19.2 billion NOK.

Also the number of participants has increased over the years. In 1996, the exchange had slightly fewer than 120 participants. In 1997, the number was almost 200, and at the end of 1998, Nord Pool had 258 participants. Of these, 95% were Norwegian, Swedish or Finnish. The rest were Danish or British. The participants are mostly divided into producers, distributors, industrial enterprises, large consumers, brokers and traders. At the market, they can act as traders, brokers, arbitrators and hedgers.

Also thepower transactions have increased, especially in the financial market. In 1996, the spot and the financial market together had a turnover of almost 84 TWh (almost 41 TWh spot), and in 1997, the figures were almost 98 MWh (44 TWh spot), which corresponds to an increase of almost 17%. But, the figures for 1998 is even higher; the physical trade increased with around 27% (from 44 to 56.3 TWh). The increase of financial contracting was even larger; namely 68%, which resulted in 91.1 TWh. According to Rickard Nilsson, Nord Pool, the large increase mainly depends on the increased number of actors, lower prices and the integration of the Finnish market.

8 1 NOK corresponded in February 19 to 7.7843 US$. 7 On 15 June 1998, Svenska Kraftnat bought 50% of the joint stock in Elex, the Finnish power exchange. Thereby, Nord Pool ’s and Elex’s actors got the possibility of trading at each others’ markets. Generally, the Finnish trade on the Swedish and Norwegian market is larger than the Swedish and Norwegian trade in Finland.

2.2.1 Physical exchange To take part in thephysical exchange, the companies are obliged to control and keep the balance between theirsupply and demand of power; a so-called ’’balance-responsibility” is required. This means that the companies have to make sure that their supply is as large as their sales.

The physical exchange is divided into a Day-Exchange (the spot market) and a Regulating Exchange. At the spot market, hourly electricity contracts are traded for the following day-and-night 24 hours. The day before, the concerned companies deliver their power balance, i.e. their supply and demand according to the spot price variations, which after the Nord Pool price calculations, results in 24 different spot prices, one for each hour, during the next day and night. This form of trade is referred to as trading in the market equilibrium, auction trading and simultaneous price fixing. The price- mechanism is aimed at regulating the power flow at limited capacity in thecentral grid.

Due to the transfer limitations, there are separate price-areas with individual prices in the Nordic market, namely the Sweden price, the Norway price(s), the Finland price and also the system price. Even within the countries, bottlenecks in the distribution grid limit the transfer capacity, and thereby, cause different prices. To prevent such differences within Sweden, Svenska Kraftnat purchases power according to a counter purchase principle 9 . To that extent, Svenska Kraftnat may also limit trade, for example between Sweden and Germany and Denmark in order to keep stable transmissions when the electricity production in the north is high compared to the production in the south.

The Regulating Exchange is an hour-to-hour exchange and handles the necessary adjustments when the situation does not follow the supply and demand plan. Big price variations may occur, for example, at a sudden drop of power production. A requirement for taking part in this trade is the ability to deliver power rapidly, 15 minutes are the maximum time. Thus, ordinary consumers have no access to and are seldom affected by this trade.

Further information regarding spot pricing and the regulating power market is given in theenclosed Paper number III.

The physical exchange also consists of a third part, namely the Balance-Adjustment. This is a kind of readjustment where the companies that did not succeed in keeping the balance economically are held accountable for their failure.

9 The counter purchase principle means that Svenska Kraftnat pays for the downward regulation of the surplus area and for the upward regulation of the deficit area. The cost connected with the counter purchase is regained through tariffs for power transmission. 8 2.2.2 Financial Exchange The financial exchange is a so-called ’’futures-exchange ” where weekly, block or season contracts are agreed upon to the spot price, depending on the actors ’ expectations of the price trend. Mainly base power is sold, which means a certain power level for a limited time. Other contract structures are also practised, as well as purchase and sales options.

The financial trade requires no balance-responsibility, and the market actors take other types of roles than at the physical market, namely like power brokers, traders, hedgers and arbitrators. The main difference among these categories is that the broker only finalizes the affair, taking little financial risks of his own. The traders (for example, banks) speculate on the price development and ’’survives ” through better price estimations than other actors. Hedgers, though, treat electricity trade like any trade of securities, and strive to secure their actions, for example, a producer might like to secure its income from future power production. Finally, the arbitrators make business through taking advantage of price miscalculations on themarket.

Further, regarding power exchange, the new risk situation discussed in section 2.2 accentuate the sensitivity of the balance between electricity sales contracts and financial contracts.

2.3 Production structure and determination of electricity prices The Swedish - Norwegian electricity production system is mainly based on hydro power (70 %) and secondly on nuclear power (26 %). Thermal power only contributes approximately 4%10. This implies that the Swedish-Norwegian power system, in contrast to most European systems, is energy-dimensioned instead of power- dimensioned. In comparison, the countries of the UCPTE11 together have more than 44% fossil-fuelled thermal power production capacity, 37.8% nuclear power, 16% hydro power and 1.9% other energy sources 12.

The variable costs for electricity production in hydro and nuclear power plants in Sweden and Norway are lower than the corresponding cost for thermal power production. The power-dimensioned systems of continental Europe to a larger extent are based on thermal power, and this partly explains why operation costs and electricity prices in Sweden-Norway are the lowest in Europe 13.

10 The figures correspond to the production conditions in 1997. 11 The Union for the Coordination of Production and Transmission of Electricity, with its members: Belgium, Germany, Spain, France, Greece, Italy, Slovenia, Croatia, Bosnia-Herzegovinia, other regions of former Yugoslavia, Luxemburg, The Netherlands, Austria, Portugal and Switzerland. 12 The figures are from 1995. " NUTEK, Elmarknadema i Europa, page 19. 9 A general rule of production optimization is that plants with lower variable costs are taken into operation before plants with higher variable costs. Regulation power, like hydro and thermal power plants, are utilized to cover middle-range and peak demand. The picture 2.3 shows the production demand and corresponding variable costs within the Nordic countries according to Sydkraft.

Ore/kWh ^ bbrmalt voter supply and availability of nuclear power 40

30

1 Import 20 power , Import 10 Hydropower

0 ► TWh/AR A 100 200 300 400 Electricity Production

CHP = Combined bEat and Potter BP = BackPiessure Plant GT=Gas Turbine

Figure. 2.3 Operation of Nordic electricity production14

The year 1996 was characterized by low water levels of the water storage of the hydro power plants. In combination with a cold winter, this led to an overall lack of energy in the Nordic system. Therefore, the average electricity price level of 1996 was higher than in 1997, and the price variations were smaller.

Because of the special condition of 1996 and because the effects from the deregulation still had not penetrated the electricity pricing, the spot prices of 1997 are chosen for the energy system analyses referred in the enclosed Paper III.

According to Hirshleifer [33], even highly imperfect markets may lead to results very close to a ’’competitive ideal ”. Thus, as the immature electricity market evolves, it can be expected that the differences between the average exchange spot price and the price from the electricity sales companies to their customers will be reduced. Real time price offers is one such example. Regarding traditional electricity sales, the sales companies ’

M 1 US$ corresponded in February 19 to 7.9585 SEK. 10 determination of prices will in the long run also become similar to those of the financial contracts at the exchange market. For instance, the power-related charges of the industrial electricity sales contracts are removed, since there are no power-related charges at the exchange market. Many customers; mainly industries, that do not have their own access to the trade at Nord Pool, have contracts with their supplier, where the pricing is connected to the daily spot price. Thereby, the determination of prices at the electricity market slowly adjusts to the exchange level.

In February 1999, the first step will be taken in the process of forming a common European electricity market. But it is likely to take several years before the separate markets are functioning as one unit. According to the institute of energy economics, Colonge University, the electricity prices within the European Union will fall at least 10% when theelectricity and gas markets are deregulated. This has already happened to electricity prices in California (10% for all residential customers) and in the UK. Studying the impact of deregulation on other product and service markets, one notes the general consequence that some market segments of each market have obtained lower prices.

In predicting the electricity price levels of a deregulated, united European electricity market, some general assumptions may be made:

• Due to deregulation the price of electricity will become similar to short-range marginal production costs [2].

• The introduction of new technology will in the long term contribute to lower electricity prices.

• When the electricity markets of southern Europe are deregulated, the electricity prices will show a downward tendency as they have done in Norway, Sweden, the UK and California [7], [23].

• Provided there are no transfer limitations and legal restrictions, the electricity prices will be similar in all Europe.

• As the variable cost for electricity production is higher in continental Europe than in Sweden in general, the price level in Sweden will raise. A parallel comparison can be made of UK gas price trends [4],

• The exchange price level, should though not in long term exceed the long-range marginal cost for new electricity production capacity (LRMC). Instead, it varies with the operational short range marginal cost (SRMC). But, the price level is likely to exceed LRMC, shortly before introduction of new production capacity. Thereafter it will probably sink back to SRMC, (except for a small margin for organizational costs) 15.

15 In theory, the price level of an infinite system, would be similar to LRMC, since introduction of new capacity would be fairly noticeable.

11

tf-. !'

Studying the price trend at Nord Pool, one sees its most characteristic attribute as the big price variations, both seasonal and during the day and night. A probable development of a common European electricity market, would though, due to different consumption patterns, be smaller variations during the day, but larger variations between the day and night. The electricity demand of middle and southern Europe differ from the Swedish curves mainly due to Swedish electrical heating. In Sweden, the load curves are relatively smooth during the day and night. In contrast, there are large differences of load level during winter and summer.

In the UCPTE countries, there are in general big differences between the load levels of the day and night, while the seasonal differences are not as large percentage-wise as those in Sweden [30]. Thus, European influence will probably contribute to increased price differences during the day and night, and fewer seasonal-bound variations. 16

SEK/MWh

Figure 2.4 Price variationsof the Nordic spot market for electricity, week 7 199916

16 Source: Nord Pool 12 Due to restrictions and the still immature market, the level of these spot prices have not penetrated the electricity prices for residential customers. But ’’lower priced ” long-term contracts are beginning to occur on the market. Though, the levels are still high compared to the spot prices of 1998. Thus, even lower electricity prices may be expected in Sweden when new actors probably confront the residential customers as the metering requirement ceases. Regarding the price elasticityof electricity, defined as ’’the proportional change in the quantity purchased divided by the proportional change in price” [33], different reports point at different elasticities of demand and cross ­ elasticities to other fuels [29]. However, also the expectations of an eventual negative price elasticity of electricity contribute to the lower electricity prices that already are offered to larger industries [81]. Examples of approximately 30% reduction are to be seen.

In contrary, the influence from European electricity prices is in the long term likely to raise the Swedish price level, as soon as political and market restrictions are dissolved.

Spot prices 1995 - 98 weekly average,

SEK/MWh, Swedish price area

During 1995 trade In Norway only

Figure 2.5 Spot price variations 1995 - 1998

2.4 Strategies for electricity sales Some major tendencies can be pointed out following in the footsteps of deregulation. Companies choose between the focus on a (low) price strategy or on value added services. Mixtures between the directions are also seen. Deciding on the first alternative, the company organization has to be adjusted (rationalized) to the new situation. Still they would run the risk of becoming an unnecessary intermediary as Internet trade expands. If not providing any extra service or pricing ennoblement, in the long run, most 13 sales companies may more or less be seen only as an extra cost for the customers. The pricing ennoblement and risk management corresponding to the trade at the power exchange might as well be performed by very few persons supported by computers, instead of by an entire sales organization. Referring to other deregulated markets, like the banking, insurance and railway industry, the flurry of price offers, interests and other privileges often seems to intend to confuse the customers rather than to bring special benefits. It is likely that the same development will be seen at the electricity market. Examples of companies concentrating on electricity pricing are Telge Energi and Agrokraft. Agrokraft also directs offers to the farming industry.

The second strategy is sometimes seen as a possibility for staying competitive also with slightly higher price offers. Most larger electricity sales companies in Sweden have so far chosen this business line, namely, to focus on VAS. The services are aimed to tie the customer tighter to the company. In contrast to a low-price strategy, which probably requires a small and flexible organization with low costs, the added value philosophy is likely to demand more personnel, customer support and some kind of control over product development. VAS are to a large extent developed to fit a certain customer group, and a primary target market seems to be the middle market segment, office buildings and complexes, institutions and retail chains.

Since the sales companies are no longer bound to the concession area of a certain distributor, differentiation on certain customer groups is a consequence. In 1997, 30% of the marketed electricity was sold off the concession, the trade at the power exchange excluded. In an EPRI17 interview of eight leaders of American power businesses [71], customer segmentation and increased diversification of the power industry is forecasted. Swedish examples are Stockholm Energi and Sydkraft, who both seem to favour estates and grocery chains. Regarding estates, Sydkraft here offers to take care of all service connected to the of the buildings, and different approaches are taken depending on theconditions.

Also the value added services are directed towards certain customer groups. Regarding their development, it can be divided in three main different strategies:

• Climbing the value chain - ennoblement • Internal competence focus • Building value networks

2.4.1 Climbing the value chain - ennoblement Regarding expansion strategies, market schools often recommend climbing the value chain [62]. The problem with electricity is that it is relatively hard to refine; 400kV is even less attractive to the customers than 220 kV. Thereby, the power industry has to look at new services, as well as market segments and market roles. Services, like light, heat, cold, climate control, 21 degrees in-house temperature, 3 degrees milk, etc., are examples of first step climbing. A second step, but also built on alliances, is the selling of electricity consuming equipment with the electricity consumption included. For

17 The American utility-supported Electric Power Research Institute 14 example, Vattenfall in cooperation with Electrolux wishes to sell dishwashers with ten years’ supply of electricity.

2.4.2 Internal competence focus Another main rule for business expansion is to focus on the internal competence capital. Looking at the power industry, competence includes not only how to run an energy system most effectively, but also the knowledge about the customers, their habits and preferences. Examples of services, where energy groups have utilized their own assets, are communication services (using existing or bought infrastructure and the knowledge of distribution maintenance) and warning of lightning (normally utilized in distribution supervision). The main driving forces behind the development are competence, customer base, IT and net structure [87].

2.4.3 Building value networks Alliances across traditional business lines are one way to create new market opportunities, since it is hard to strategically renew an established company. Weiner [83] describes the function of the value network as when ” one company exploits the strengths of each value provider and coordinates production and delivery across companies. The leader in a value network coordinates the activities of other companies in the network, choosing and assembling the capabilities to deliver value to a specific customer segment. ” He continues, ’’By not owning all the assets in the value chain, the company is a virtual organization. ”

In any utility value network, the players will use different governance mechanisms, such as, partnerships, alliances, joint ventures, specialized contracts and outsourcing arrangements to manage their relationships. The Stockholm Energi cooperation with Glocal Net for cheaper telephone services for their customers is one Swedish example. According to Tord Ingvarsson, project manager of Gotcom (see section 3.5.2), Vattenfall has the intention of building alliances with other manufacturing industries to become a full-service company [22]. In the UK, large alliance-built bonus systems are alreadydeveloped.

Also under way are pilots to bundle electricity with other utilities, such as, gas, telephone and cable. To broaden the perspective, it is also likely that the future power industry will consist of a mixture of energy trade by addition of gas and oil. Jan C Johansson, Vattenfall, forecasts in an interview [22] that within ten years the oil and power businesses will be integrated to one and the same business. Already, Gullspang and Preem are examining the possibilities for a joint energy sales company, Norsk Hydro sells oil, petrol and electricity to their customers and Nord Pool has formed a strategic alliance with IPE18 in London. A united project between these two companies aims at simplifying cross-trading for their respective customers, i. e. to give them the possibility of purchasing both oil and electricity.

18 International Petroleum Exchange 15 3. I.T. IN THE POWER INDUSTRY The intention of this chapter is to provide a picture of how IT currently is utilized within the power industry, power generation excluded. It also aims to show how the deregulation already has and will continue to affect future IT application within the power business. The main emphasis is laid upon sales-related IT applications, but due to the interdependence of thenetwork-operators and the electricity sales companies (ESC), IT-based applications of the NOs are briefed as well. The overview concentrates on a regional and local level, therefore, IT applications of a national level, for example, by Svenska Kraftnat, are not included.

The applications are divided in four areas: IT applications in power system balancing, metering, distribution maintenance and applications in electricity sales. The main function of each area is first discussed in terms of its objective and IT utilization, and secondly, in regard to future IT applications and driving forces. The chapter ends with a short description of power-related IT projects.

The observations are based on interviews, on research reports and on articles of business-specific journals. Literature surveys gave no references regarding the perspective of ”IT utilization in power industry and electricity sales” as a whole. Sources referred to are, instead, specialized within a certain area, like distribution supervision, DSM, etc.

3.1 IT applications in power system balancing IT applications for maintaining the physical balance of the power system are here divided into the two areas of production optimization and distribution supervision. The picture 3.1 show the main information flows that ensure a regional power system balance.

16 POWER PLANTS

BILATERAL CONTRACTS WITH OTHER POWER SUPPLY COMPANIES LARGE

CUSTOMERS POWER

PRODUCTION PLANNING NORDPOOL POWER EXCHANGE AND CONTINUOUS

PRODUCTION OPTIMISATION

ELECTRICITY

SVENSKA KRAFTNAT SALES COMPANY

DEDUCTION

DISTRIBUTION DISTRIBUTION SUPERVISION CONCESSION AREAS

Figure 3.1 The main information flows required to ensure a regional power system balance

Concentrating on the function of production planning and optimization, briefly, the following information is exchanged with its surrounding units:

1. Availability of production capacity and requests of operation.

2. Customer production plans and eventual requests for disconnection of load.

3. Information about contracted deliveries, capacities and prices.

4. Deduction of customers.

5. Metered data of consumption and climate key-factors.

6. Data for long and short term production optimization, power balances and capacity utilization, physical trade etc.

7. Eventual status reports and requests of certain power balancing measures.

8. Measuring series for the deduction.

9. Deduction check

10. Deduction data

11. Short term production capacity and requests! of certain power balancing measures, offers for the regulating market.

17 12. Information regarding physical and financial trade.

13. Agreements and requests.

3.1.1 Production optimization Production planning and production optimization are performed to secure the supply company’s committed electricity delivery and to minimize the costs. The fundamental principle is to coordinate the company’s own production capacity, estimation of demand and the expected power exchange. Efficient electricity deduction systems and reliable metering are required to sort financial interactions among different actors. To succeed, advanced IT-tools are utilized.

To estimate the short- (and long-) term demand, real time metered data of consumption, temperature, wind force and sun radiation, at certain measuring points in the grid, are continually, via the deduction, provided to the production planning system. The objective of the deduction is, in this case, to, based on earlier experiences, create reference data of power balance according to the metered consumption. Thereafter, these data are in modem control systems automatically imported to the database of the production panning and coordination.

Based on these figures, namely the reference consumption values from the deduction, the computer system record the power balance every 90:th second. According to Forsberg, Sydkraft Produktion, longer time periods between balance control are also possible, as long as the deduction of power costs is only performed by the hour. This is also probably a practical solution, but theoretically, it obstructs energy system optimization. On the contrary, for optimal control, real time is required. In the USA, normal cycles for data collection are 10-12 seconds [52].

For the power balance deduction in Sydkraft, each day after the deregulation, reports from around 375 of the 450 grid areas and the deduction from around 250 grid owners, are handled. Defined measuring series for every net area together brings an aggregated sum of all hourly metered customers. Added to an unspecified part, an interruptible part and the actual production, these figures constitute the base where upon the economic deduction is performed. Sydkraft receives around 1400 measuring series for 24 hours back in time each day. This is to be compared with approximately 150 reports before the deregulation. The information is sent in various ways, by fax over telephone lines, or electronically on the protocols of Svenska Kraftnat.

Against the electricity price at the exchange, the calculated consumption patterns and the financial and physical agreements of the power exchange, the generation of the production plants are manually optimized.

18 Regarding physical power trade, through the estimations of the, based on experience, coming day price variations, and depending on the supply of regulate hydro power, the trading strategy and the physical trade scheme committed to Nord Pool is formed.

In the bid, consideration is taken to eventual changes from the estimated prices. At higher spot prices than estimated, the purchases decrease or the sales increase, and vice versa at lower prices.

Ntfce 50 too 150 200 Conutted power bid [MW] 1 1 X" y Estimated spot price variations 2 -100 150 Vo 2501

3 * 300 t\ 3V 4 250 1 / /

Figure 3.2 Protocol ofplanned power trade

In essence the strategy is to dispose capacity to sell when price and consumption are high and to buy at times of lower prices and demand. For example, at 02 am, provided that the spot price is 150 SEK/MWh,the company commits to sell 200 MW. If the price instead is 200 SEK/MWh the company commits to sell 250 MW. In contrary, if the spot price is 50 SEK/MW the company commits to purchase 100 MW.

Status information of the generation units is electronically provided the to control room of the coordinating production company, either for supervision only or for control. For example, hydro-power plants are advantageously controlled remotely. In contrast, power requests of nuclear power or fossil-fuelled condense power are usually made by telephone at least one week in advance. Also, situations very often occur which require daily corrections of previous plans.

Functions within traditional SCADA19 systems are often utilized for the performance of theabove-mentioned activities.

19 Supervisory Control and Data Acquisition 19 Regarding the physical power exchange with Nord Pool and Svenska Kra final, the information and the bids of power are exchanged on Nord Pool specified protocols via Internet (ELWEB - new system) or fax (old). The communication with Svenska Kraftnat, performed via fax or electronically by EDIEL protocols, mainly concerns questions of distribution stabilization by power exchange at the hourly Regulate Exchange market ( see 2.3.1) Other agreements of importance are bilateral contracts signed with other electricity producers. All information about power trade relevant for the production optimization is so far manually tapped into the computer database. This condition is changed with the introduction of more up-to-date control systems.

Other processes of importance for the financial interactions are company deduction, which is the base for charge, transaction deduction and compilation of statistics. Transaction and company deductions are performed and communicated (in paper form) to all parties concerned. The communication with Svenska Kraftnat is electronic and is standardized to their protocols. Compilation of statistics for internal long-term analyses are produced in paper form.

3.1.1.1 Outlook Before the deregulation, load estimation and deduction was relatively easily performed, since the metered supply in the supply point of the concession area minus the estimated losses, expressed in percentages, also represented the actual electricity delivery. This was a consequence of having all the customers sorted under the company’s own distribution grid. On the free market, former customers have chosen to change supplier, and new customers are won outside the area covered of the company’s own distribution. Therefore the estimation of the required and also the final deduction have become more complex, since more measuring points are required, more actors are involved and thereby more data need to be processed due to the deregulation. Sometimes the metering is even still performed in the same metering points. The picture 3.3 shows the new complexity of themarket.

20 _ _ .CONCESSION AREA 1 CONCESSION AREA 2

DNOl ESC = Electricity Sales Company

Figure 3.3 Several Electricity Sales Companies may operate within the same distribution concessionarea

Generally, the new complexity of the electricity market requires enhanced communication among its actors. As customers are won in new distribution concession areas, the information to process increases together with the requirements of efficient electronic communication and direct data import to powerful deduction and optimization systems.

Former computer systems are seldom built to handle all this information; thus, change to new, more open and compatible systems often is required [68].

The above-mentioned deduction model is an easier alternative to the sometimes recommended on-line estimation of customer loads, a method which would require even more IT-supported applications. Such a model would require hourly past recordings of all customers, customer class models and on-line measurements in the distribution network [50]; i. e. activities that especially would make heavy demands upon the distributors.

Dahlfors and Pilling proclaim that information management will become one of the most strategic resources for a power company seeking to be successful in a competitive market [18].Systems and communication techniques that enable efficient and directed information distribution will, therefore, in the future become one of the company’s most valuable assets. The three major reasons are 1) the new role of the energy utility (see 2.2), 2) the horizontal and vertical integration (like integration of administrative computing with DSM20 information and control centre with local protection systems) and 3) open IT industry standard models.

Not only load estimation and deduction will be affected by deregulation and the IT-era, but also, power system operation and market applications will become integrated to new Business Management Systems (BMS) [18], [52], which, according to Pilling, is a necessity for the producers to stay competitive. This is a relatively new area of system integration for energy trading and customer service, which links the power system operation with its business-related processes mainly to obtain a support system for simplified strategic market decisions and spot market trade. Hereby, for example, the creation of the protocol for planned power trade (see fig. 3.2) could be automatized.

For efficient utilization, the BMS are likely to require standardized and direct electronic data import from the power exchange. Manual operations will be avoided to the greatest extent possible to reduce extra costs and to increase the efficiency of the systems. A consequence is that the optimization systems of the past seldom will be powerful enough to process all the new information, or their structure will be incompatible, so that these systems will also need to be exchanged.

Other activities that can be augmented with the help of IT applications are utilization of distributed resources and load control for cost optimization of energy systems. In the case of limitations in distribution capacity or in central plant power generation, distributed production capacity can be the only solution to maintain reliability, or to avoid heavy investment costs [67]. In Japan, where location of large-scale power plants is a severe problem, the information systems for distribution supervision have to be designed to optimize also the utilization of load management and distributed resources [57].

Farming industry, hospitals or other industry often own distributed resources that can be utilized by the power industry, for example, to alleviate power shortages or to make capacity available at the power exchange and for minimization of the energy system cost as shown by Andersson [19], [2],

Although one-way communication theoretically could enable instant activation of distributed resources for system optimization, or for the single benefit of the exchange actor, two-way communication is probably the most practical solution. This is a conclusion already drawn by the operators of the distribution control and supervision. One direction is required to inform of equipment status, price of activation and eventual operation plans, and the other to accept the terms and to control the equipment.

Another possibility is, of course, thatthe concerned businesses themselves would prefer to trade at the power exchange. This development would also require IT solutions (like Internet communication) to provide all the actors of the exchange with exact and timely information.

Concerning DSM, according to Lehtonen [50] the increased (on-line) metering, the associated data transmission capabilities and the enhanced freedom in agreements

20 Demand Side Management 22 between customers and their suppliers caused by the deregulation will open new possibilities for DSM actions. Direct load control will permit, in periods of high prices, the actors of the power exchange to increase their power supply that can be sold at the exchange. Thereby, their assets due to IT solutions would be utilized with larger economic benefit.

Dahlfors and Pilling [18] mean that commercial management of the supply/demand balance against flexible price structures and themaximized utilization of low price base­ load generation, will become necessary for successful competition in the supply market.

Mostly larger loads are discussed, but jointly the aggregated sum of loads controllable in the residential sector are of as much importance. Large-scale implementation of direct connection for small customers to the power exchange would require new IT functions for hourly metering, deduction and eventual load control.

As for distributed resources, the major benefits from DSM depend on the optimization of the energy system assets. During theyears, many DSM projects have been performed [19], [77], [3], with more or less successful results regarding system cost savings. As the deregulation allows the short-term marginal cost to become evident through the exchange market, more flexible, continuous and automatic solutions are also required to respond to the market conditions of each hour.

Protocols are being evaluated for DSB (Demand Side Bidding) systems. Here, the costs of further production capacity are weighed against the costs for load control. The main problem so far is how to secure fair market conditions for all actors, i. e. how to avoid having the customers set their interrupt price depending on the price trend at the power exchange.

3.1.2 Distribution supervision In electricity distribution, IT is utilized in two main areas:

• in operation for supervision and control of distribution network balance, metering and administrative data transfer

• in maintenance for increased efficiency, etc.

The main tasks for the distribution supervision are to balance the electricity grid to the greatest extent possible, to secure the reliability of the delivery, to plan for maintenance interruptions and to minimize the interruption time. The overall aim is economic optimization of disposable assets against the constraints of reliability. This is achieved, among other ways, by momentary control of certain objects and metering at specific nodes in the distribution grid. Both functions require IT solutions, one- way to collect data and two-way to perform the required switching, etc.

The data transfer aims to provide the grid operators with the current status of selected measuring points of the grid. Based on this information, for example, fault diagnostics and reliability analyses are performed. Traditional SCADA systems are utilized for the control, supervision, storage and collection of the power system information. 23 Transmitted power, reactive power, current and voltage are the main parameters measured in real time. Alarm functions are built in the control centre system to warn when certain parameters risk to exceed their respective limit of capacity.

Based on the status of the grid, the operator decides upon certain measures, such as, switching for fault isolation, service restoration and/or maintenance, voltage regulation, and losses minimization, for example, through run of gas turbines for power correcting, frequency regulation, etc. In Sweden, the latter is often automatically obtained by automatic frequency regulators at the turbines of the hydro-power plants. Performing most of these operations, the operator gets little support during severe situations by the traditional systems. Network maps and intuition in switching planning are sometimes complemented with results from network calculating programs.

The weakness is that these programs seldom are constructed for real time calculations [82]. The gravity of the situation and what measures to perform are, instead, evaluated by the operator himself based on his experience. Centralization of control centre areas aggravates the situation. For example, in 1992, Sydkraft had 16 manned operation units, today, there are only four, mainly due to economic benefits from rationalization. Increased IT utilization is a necessity to enable the centralization and corresponding rationalizations.

To increase the reliability and to avoid unnecessary losses, Distributed Automation (DA) is further utilized within distribution supervision and integrated with the SCADA system. For example at Sydkraft all regional net stations and most of the distribution stations are equipped with remote control and from these stations measurement results and relay indications are transferred to the operation central.

Energy Management systems (EMS) and/or Distribution Management Systems (DMS) for integration of production optimization and distribution supervision are lately often added to the SCADA system [82], [55], [68], [52], [54], [28]. They are generally described as an expansion of the former SCADA and aim to provide supporting software tools for control centre operations. EMS/DMS contain functions for production planning and control, network configuration, topology supervision, network and reliability analyses, etc. But at large-scale disturbances, neither these systems manages to support the operator, on account of large quantities of data [52].

One solution for reducing the operator intervention for switching purposes together with line losses and energy consumption is to place intelligent functions distributed in the network. Southern California Edison Company has developed an automated capacitor switching system for voltage regulation, which directly undertakes required switching, optimal for the entire system. The capacitors use new electronic meters to register real­ time data of customer voltages and energy consumption. By two-way packet radios, the data is communicated to the central data system of the company where a control algorithm determines the optimal capacitor switching [85]. The result is thereafter communicated back to the capacitors. Test results indicate that, except for the reduced amount of central operations, about 1% energy savings are gained for every 1% of voltage reduction.

24 3.1.2.1 Outlook The development of IT and the increasing importance of distribution system operation has led to further utilization of DA [82]. Even though the distribution has stayed in a monopoly, distributors are still interested in efficiency improvements. According to Persson, in Sweden, there are two main reasons; first, the Board of Supervision of electricity distribution aims at increasing the efficiency of the power industry, and secondly, it is still to some extent possible to earn money from improved efficiency. Thirdly, the authors remark, the costs of electricity distribution have become more evident after the deregulation due to the split of electricity sales and electricity distribution prices on the customer tariffs. Thereby, the deregulation, except for the hourly meter requirement (see 3.2), also indirectly has affected the NOs.

A probable consequence is, therefore, increased utilization of SCADA and EMS/DMS also by smaller distributors [52]; i. e. IT will be further utilized also to optimize the network operation. Kam Hong Cheong [17] explains the phenomenon of the fast growing area of Distribution Management as a result of three separate driving forces: 1) advances in technology, namely microprocessors, powerful communication links and sensors, 2) the deregulation and the new role of the utilities and 3) the increased sophistication of the society, which has become more vulnerable to disturbances.

Based on experiences at Malmo Energi, Persson ascertained that higher degrees of automation in the distribution network will lower the cost for operation and increase the power quality from the customer ’s point of view [70]. However, the costs for the DA equipment are still high; therefore, the utilization of distribution automation has partially ceased.

Looking at the distribution system of North America, which is among those considered to be thelargest operating machines in the world, due to economic pretensions forced by the coming deregulation, reinforcement is put on hold. According to Lof [52], instead companies strive for increased utilization of all power system assets. In California, this has led to an increased transmission of electric power to large load areas from distantly located production plants. Thereby, the transmission grid is operated on the edge of its physical limits. To prevent large-scale failures, IT has to be utilized to enable optimized network operation.

A probable consequence of the Californian situation is increased utilization of DSM; in accordance with thetheories of resource economizing, that utilizes the existing assets to the most extent possible. According to Lethonen [49], themain economic benefit of load control for the distributor lies in saved investment costs in extended distribution capacity.

The enclosed Paper II, though, shows that Swedish distributors have few incentives for load management after the deregulation. As relatively few bottlenecks are to be found in the Swedish distribution grid (due to reinforcements in the 1970s caused by forecasts of large electric heating demand), load management often can be placed side by side with a direct loss of income for these distributors.

The incentives for load management have, thereby, left the distributor for the ESC, the customer and, sometimes, also the producer (see 3.1.1.1). The main reason for 25 distributors to introduce load management would be to anticipate the sales company and to adjust the load control to theirown most preferable conditions.

Introduction of FACTS (Flexible AC Transmission systems) is an additional area under development, where IT is likely to contribute to increased transmission capacity of the grid. FACTS enables fast, direct control of equipment, either automatically or by the operator, and can thereby control power flows, reduce transmission losses and increase the flexibility of the power system. The components are mainly based on thyristors for fast variation of voltage, current, the phase angle or impedance. The utilization of FACTS is forecasted to shortly change the power systems from passive energy transport networks to systems where powerflows can be affected automatically or by the operating engineer. Corresponding FACTS components are developed for both low and medium voltage distribution.

An interesting point is the potential competition between FACTS and distributed generation. As FACTS and other power electronics make the power delivery system much more flexible and efficient, distributed generation might take the customer off the grid entirely.

Development of intelligent agents for distribution supervision is another trend trying to improve the efficiency of the power grid without deteriorating the reliability. The idea is to allocate intelligence in the grid to improve the efficiency and to optimize the distribution operation. In the vision of Wildberger [84], who creates simulation models of intelligent agent based distributed control, locally operating intelligent agents (IA) will automatically perform local control based on the optimization of the entire power system. By using a combination of genetic algorithms and genetic programming, these agents will evolve by active learning in real situations. The approach places automated decision-making and computation at the same location in the grid as are the sensors and controllers. Through two-way communication with other equipment, like FACTS, etc., the optimization of the entire system will be performed.

3.2 Distribution maintenance The maintenance within the electricity distribution has a history of a "run to failure” approach via planned periodical service to condition-based maintenance. Even though condition-based maintenance is an established conception for many years, many companies still keep to the periodical service. For instance, in Sydkraft, this means that the distribution network (0.4-130kV) is inspected manually every eight years and the transmission grid every year by flight inspections. Transformation stations are overviewed every fourth year, etc. The documentation of equipment condition, like circuit-breaking, etc., is as elementary as the maintenance methodology. It is often made in paper form as there is no maintenance database. Based on their own experience and without system support, the service operators perform the required measures. Transparencies, but no digital maps, help them to orientate.

For fault allocation, remote switchers are installed in many distribution networks. In Sydkraft and by other distributors about 20% are found to be an optimum installation amount according to corresponding costs of emergency service. The switchers have automatic fault out sectioning, which helps to locate failures and, of course, to minimize 26 the need for emergency service. Other equipment for fault positioning are accidental ground and short-circuit indicators installed at natural branching in thegrid.

Southwest Queensland South West Power, together with Introl, have developed a control system based on intelligent Programmable Logic Controllers. Among other services, it enables multiple control points to its substations and distributing operating devices. By a laptop computer connected to a modem or the mobile phone, the system can, for example, be controlled by the service operator in the field or by the support staff personnel. The methodology minimizes interruptions and increases the flexibility of operation in response to emergency conditions.

3.2.1 Outlook As the costs for electricity distribution have become evident due to the deregulation, there is increased necessity for the distribution companies to improve their level of efficiency. This also brings a more or less insolvable problem: to contain the reliability of thedelivery and reduce the costs for themaintenance.

As mentioned above, IT solutions for efficiency improvement in distribution supervision is one main strategy. Integration of maintenance databases to existing SCADA equipment is another. Several companies are evaluating Asset and Maintenance Management Systems (AMS), which link the information of the SCADA system to, for example, supply data, Geographical Information Systems (GIS), service planning systems, fault reports, network analyses, maintenance and sometimes even commercial information, etc. By direct connection by laptop to the company data system, the service operator will be provided with all the information herequires directly at his location. He will get access to work orders, which tell him his daily scheme, and to digital maps of a concerned area to simplify his mission. Central support units might assist him in complex situations. By pre-specified database protocols, he can make measure reports which are directly stored to be further utilized for maintenance planning, etc. In the AMS, both real time and historical data will become accessible, for example, for observing network coverage (occupation) and frequent surplus capacities. Tariffs might be adjusted to mirror the grid status and demand equalization achieved. This is called CMMS, Computerized Maintenance Management System.

By utilizing the information of the actual condition of certain equipment, systems for Reliability Centred Maintenance (RCM) can be applied. RCM is to be seen as a combination of condition-based, preventing and remedying maintenance. It adjusts the maintenance of the power grid to correspond to the functionality of certain equipment and the system economic consequences of a sudden failure. RCM is currently tested, among other places in the UK [68], and software is developed to support the method which aims at establishing a maintenance that concentrates on the functional effects and the historical data of failure of certain equipment, instead of focusing on the recommendations of the manufacturers. To manage all this data, new information systems are most often required.

27 3.3 Metering Today, the distribution network operators (DNO) are obliged to perform annual metering of customers within their concession area and to provide the information to the concerned electricity sales companies. Since the meters of small customers are read on the location and only once a year, it might cause considerable overcharges or underestimations. Until November 1999, due to legislation, customers that prefer a supplier other than the traditional operator, have to install hourly meters to register their consumption. The cost for such a meter is around 2500 SEK. But Vattenfall also drive the power business to disassociate it selves from the today established preliminary charges. Due to the deregulation exact charging will probably be one of the "services ” that soon is seen as a "factor of hygiene", the same way as EDI already has become (see 3.4.2).

To take an offensive position in the competition among the electricity suppliers, Vattenfall announced in autumn 1998, that all their customers would be offered an hourly electric meter, free of charge. As Vattenfall will use the meter as an entrance to the IT-home market (see 3.5.2), this can be seen as a natural step in their strategy. The customers would not change supplier, since Vattenfall would offer them extra service functions together withthe meter.

3.3.1 Outlook Based on these conditions, it is likely to be an enhanced utilization of IT for electric metering in the near future. Due to new business structures (see 2.2) and the need for interaction with the customers, Stoa recommends a two-way communication system for customer metering [78]. His argumentation is the free market demand of rapid and open information flows to all market participants. To obtain equal market conditions and optimal market balancing, offerings and sales should simultaneously be available for all actors, and in order to ensure the market cash flow settlement, data should be fast collected.

Also, according to Dahlfors and Pilling, two-way communication to each load point will be required to fulfil the objectives of efficient remote metering in a competitive market.

Different metering equipment and communication techniques are under evaluation within many companies. Remote metering via telephone, radio or on the power lines is likely to become a future business area. New specialized actors, MOs (Meter Operators) will probably take over the metering from the DNOs and serve many different distribution utilities.

3.4 IT applications in electricity sales The objective of the former market division, i. e. the electricity sales company, has not changed due to the deregulation, although their methods to reach that aim partly have. Still, the goal is to maximize the income from electricity sales. Under monopoly conditions this was achieved by maximizing the sales volume; now volume and price need to be weighed against each other to obtain the maximum margin between the incomes from electricity sales and the costs for electricity generation and purchase.

28 As discussed in section 2.4, to keep old and to win new customers, market orientation is shown by introduction of value added services (VAS) and also in increased customer support enabled by enhanced customer management. This is also the division chosen for illustration of the companies ’ utilization of IT applications. But, since, for example, a call-centre function can be seen both as a service for the customer and as a support tool for the billing system, thedelimitation between the areas is not always evident.

3.4.1 Systems for customer management This group consists of systems and communication for customer databases, call-centres and billing.

The most strategic asset of the free market is the information of the customer; his economic and technical status, namely the living conditions for residential customers, his energy consumption, namely how he uses electricity and other energy carriers, and his service preferences. All this information is stored in customer databases. The common problem is that it is not accessible for everybody concerned, since the data is most often stored in several poorly synchronized databases.

Performed interviews bring a similar picture. Before the deregulation, only small amounts of the information in the customer databases were utilized, sometimes only around 2%. An important factor was the absence of links among the databases, the system size and access difficulties. Also, the technique to sort and search information was expensive, as the systems of yesterday were not designed to meet this demand.

Therefore, according to Persson [68], to prepare for the deregulation, regional electricity utilities in the UK invested an average of roughly 870 MSEK and 500 SEK/customer in IT systems. A large part of that investment was aimed at improved systems for customer management.

Effectively utilized, the stored data could contribute with customer information, which, for example, would enable directed and tailored offerings to a specific customer or customer group. Through analyses of the data, it would also be possible to estimate the potential for an individual customer to switch supplier.

Different approaches are taken to solve this problem of information management. Either the information is centralized (see picture 3.4), or it is distributed on several linked databases (picture 3.5). Which solution is chosen is significant for most strategic and practical communication for customer management and value added services. Most important is that all persons working with a customer have easy access to all of the well- updated customer information. The same information must be accessible for all persons. Secondly, the salesperson must be able to reach the information he requires independent of his location. He shall also be able to load new data.

29 Business Business Business Business process process process process

Database

Figure 3.4 Single database approach

Due to rationalization, many companies choose to centralize the information into one database. Operation and running costs are said to become lower, and large system purchases can bring a certain cost reduction in the short term. Problems can occur though, when one wants to add new functions to the original system, and the additions require other database structures. The openness of the centralized system is often limited and the centralized information storage also makes these systems more vulnerable. Further, advanced market analyses, etc., that directly process information stored in the database, risk to delay, all other operations about to be performed in thesystem.

Linked system databases have the advantage of increased flexibility and increased openness. Automatic information upgrading can be added through ’’middleware ” to obtain the same or the relevant information for all connected systems. Immediate updating is performed directly at all locations when somebody changes the information in one database. Data warehousing can be utilized to obtain strategic customer information from all the required bases. Another advantage of data warehousing is the possibility of performing analyses without disturbance of the production level. According to the Gartner Group [75], in the future, enterprises are likely to have about the same level of hardware and software heterogeneity, but more application data and function duplication than they have today.

30 Figure 3.5 Multi database approach

Either way, advances in information technology, solve the problem of efficient communication to ’’optimize ” the customer management systems.

Billing is another area where large improvements can be performed. Not only the appearance of the tariff, which has been criticized, but also the entire billing procedure can be improved. In the UK, smart cards can be loaded or discs bought to pay in advance to an installed in-payment meter. Thereby, the power industry can avoid conflicts emerging when customers do not pay their bills.

A popular payment system is autogiro. Surveys of other retail industry have shown that customers that pay by autogiro have significantly less switching potential than customers using other systems for their payment.

Routines for payment by the Internet are also under development, mainly driven by the banking industry. Security issues are the dominating hindrances. The purchase of electricity on the Internet is alreadypossible.

Another trend is the expansion of call-centre activities. Twenty-four hour support has become a matter of course after the deregulation. The call-centres are mainly centralized and equipped with powerful IT-tools. This development can be compared to that of the power system distribution supervision (see 3.1.2.1). Their main tasks are to answer customer inquiries, and to handle mischarges and other error messages. Due to the information society, call-centres are geographically independent. Looking at other 31 industries, for example, the telephone exchange and the customers support function of IBM Europe is located in Greenock, Scotland.

According to Andersson [4], in the UK, the concentration on IT-based applications for customer information, supervision and business systems is partly a result of having halved the number employed within the power industry since theprivatization.

3.4.1.1 Outlook Improved utilization of customer information will dominate the customer management processes. A not very considered consequence of the acquisition strategy (see 2.2) of powerful actors in the deregulated market is the large number of administrative and customer management systems that are to be incorporated in the organization. Just at Sydkraft, alone there are more than ten different systems for billing and more than five market-related databases for customer information. Therefore, due to the deregulation, openness is the most strategic issue concerning future information system management. According to Lundgren, Sydkraft efforts to solve all issues in one giant system seem to be failing.

Regarding call-centres, two major trends contradict each other. Firstly, telephone exchanges with smart support systems enhance the customer service; while, secondly, the Internet might take over large parts of former call-centre activities.

To start with, the advanced telephone exchanges that are offered simplify the customer support. Functions to be obtained are, among others, immediate identification of customers by telephone number, customer priority orders, for example, mobile phones, provision of historical data, categorizing of customer matters and direct system support by links to relevant customer protocols, etc.

Most of the matters handled today by the customer service are billing-related. Many of these businesses could and can already be managed over the Internet. The customer interaction on the Internet will, therefore, reduce a large part of these call-centre activities. Thus, the pressure on the call-centres is likely to decrease, but also other, more complex matters will partly supersede the former tasks. It is most likely that the power business will follow the steps of other deregulated market participants. For example, like the insurance and banking industries, they will increase the customer support, introduce ’’personal ” advisors, market telephone service, either personal or computerized, and then enable Internet service solutions.

Another consequence of the deregulation is the demand for more modulated customer information. Shortly, it can be expressed as methods to enable directed ’’personal ” marketing and Internet interfaces to mass market participants. By documenting the response from a customer on certain issues, for example, on the Internet, methods are identified to characterize the customer by his demand for information. By using this data, offerings can be tailored to fit a certain customer [47], [48], [9],

32 3.4.2 Value added services Looking from the outside, in their creation of new attractive customer services, most Swedish electricity sales companies, seem to have made more or less similar analyses of customer demand and driving forces to focus. According to Vattenfall, the electricity supplier that liberates time, money and assets for the customers and improves their comfort and environment will also succeed in the competition for the customer. The business concept of Sydkraft adds ’’improvement of the customers ’ competitiveness ” as a further issue. The latter also brings a slight dilemma as value-added services themselves, sometimes compete at their customers ’ market [62].

Energy-related services Energy inspections X X Personal advisory service X X X Secure electricity deliveries X (X) Power demand optimization X X Improved energy efficiency X X (X) Continuous measurements and data X (X) X X presentation

Energy demand estimation X X Guaranteed savings X X Equipment control against tariff X Remote control of electricity utilization X X X Source declared power deliveries X X X Hourly metering X X Complete heat supply, process heat or X (X) ordinary heating Outsourced energy-related equipment X control

Environmental services Environmental round X Environmental training X

Environmentally declared power X X deliveries Improved environmental data X (X)

Administrative services Meter reading X X X X Give notice of changed residence X X Energy statistics X X X

Demand profiles for each bleeding point X X X

Invoice elucidation X X Optional number of invoices X X Optional invoice frequency X X Optional expiry date X X Autogiro X X X Electronic trade and EDI X X X

33 VALUE ADDED SERVICES , CUSTOMER GROUP . ; <:!T utilization .. COMPANIES RESIDENCES FOR ADMINI-OTHER STRATION

Voice answer X X

24-hour customer support X X X

Information services Energy News X X X Energy-saving tips X X X Price and tariff information X X Price comparison with other companies X X

Information of deregulation X X X Ordering of printed matter X X

Price offers Real time pricing X X X Fixed electricity price X X X Long-range agreement X X X

Other services Energy-related products X X X (X) Other products, books, etc. X X Mail-order bonus X X Club memberships X Communication services X X X X Coordination of electricians and plumbers X X

Lightning warning X X Insurance X X X Children security check X X

Figure 3.6 Table ofVAS offered on the Internet

Studying the company homepages on the Internet the list of offers in figure 3.6 can be compiled. The list is not intended to be complete, but to mirror the most common services on the market. Most of the studied companies 21 and all of the large electricity suppliers, like Sydkraft, Vattenfall, Gullspang, Graninge and Stockholm Energi, focus on energy counselling, easily accessible customer support and declared power deliveries. Many services are energy-related like, efficiency improvement analyses, energy inspections, etc. But tendencies pointing in new directions can also be seen. Of these, so far, guidance in environmental issues and the administrative Internet applications dominate the picture.

Specific examples that show that some companies reach further than their traditional area; energy, are lightning warnings (Sydkraft), communication services (among other companies Lunds Energi and Helsingborgs Energi), insurance (Stockholm Energy) and coordination of electricians and plumbers (Gullspang).

21 see web-addresses in references 34 To enable the two latter services, insurance and coordination of labour, alliances and agreements are made with partners who can provide these extra values for the customer. The energy company has here taken the role of the coordinator (see 2.4.3). This can be seen as a step from the traditional value chain (see 2.4.1) to new more network- orientated services.

In the two first services, by among other companies Sydkraft, Lunds Energi and Helsingborg, the companies have focused on the values of their own assets, network and competence capital, and how these can be utilized to provide new customer services. Theoretically, this is a strategic rule of how to expand and ennoble a company’s business operation in times of continuous change (see 2.4.2).

Warning of lightning is an established concept within the power system distribution supervision, information that easily can be transferred to interested industries. Regarding communication services, over the years the power industry has built up an extensive infrastructure, mainly to meet their own communication requirements. In the USA, the electric utility industry is ranked second in its use of telecommunication media 22. According to Lamarre [45], only about 3% of the communication capacity is utilized though. By making use of already-made investments, the electricity grid or other communication assets, the power industry can offer the customers new VAS. Operation and maintenance of distribution channels are also relatively well-known experiences of the NO. Thereby, communication is a logic increasing service as the power industry expands its operative business area or brings VAS to their customers.

Concerning the communication infrastructure, within the power industry, the distributors are the most natural operators. Swedish energy utilities have to a greater extent, also joined the Swedish Urban Network Association, a non-profit-making association for extension of IT-infrastructure, open for all actors that own and operate communication networks available for all individuals. Typical members are cable-TV companies, municipalities and electricity distribution utilities.

Looking at theInternet, most companies, both regional actors and local utilities, provide Internet homepages although, the quality of the web-sites differs. This goes for both the NOs and the ESCs. The customer interaction at the sites is limited and mainly extends to registration of consumption data or registration of changed residence. Ordering of administrative services and certain products is also possible. Many of these products are later sent in handwritten envelopes. In general, the larger energy corporations also provide more interactive web-sites, while smaller utilities mainly assist with company and price information.

Another type of interaction is when customers can compare their energy consumption and price to figures calculated by a residence-consumption program on the Internet. At the end of the program, he can contact the supplier for electricity ordering or for more detailed analyses. Different variations on this theme are provided by at least four Swedish sales companies.

22 The communication industry itself ranks first. 35 In some utility pages, services provided by their main contracted supplier can be recognized. Thereby, these services also have fulfilled their task of tying the customer, i. e. the local energy utility, tighter to the supplier (see 2.4)

It can be discussed whether some of the administrative functions provided on the Internet really are value added services or rather simple administrative applications expected by the customers. According to Paul Lundgren, Sydkraft, examples of such functions are EDI payment and compiled invoices. It should be a matter of course; ”a factor of hygiene”, to offer the customers these services. The author’s belief is that the same development will be seen concerning exact invoices. According to Augustsson [7], preliminary charging would never be accepted by American customers.

Around 90% of the documented services utilize IT, mainly for administrative reasons (70%). Most of these services are also available by ordinary contact, such as phone, visit, regular mail, etc. IT applications in technical solutions stand for around 30%23.

Looking at the Internet marketed services of some of the energy utilities in the UK and marketed services in the USA according to Swedish Technical attaches [23] and Goldman [27], many of these are similar to those documented above, but some interesting exceptions are also noticed. In the UK, offers of energy-related services are not provided as often as they are in Sweden. The reason why is the belief that advantages in the competition for the customer will not be won in the energy sector, but rather in areas of larger interest for the customer. Thereby, these are the development of multi-utilities and the emergence of new cross-business alliances.

The concept of the ’’multi-utility” is also, if not established so at least further developed in the UK than in Sweden. Multi-utilities tend to add the supply of water, other energy carriers, like gas and oil, and telecommunication to their service flora.

Of the companies focused on strategic alliances, most successful are different bonus systems, based on credit cards. At least two of the three dominating actors 24 of the British market, namely (SP) and British Gas (BG) engage in such concepts. SP has launched a VISA card with loyalty ’’Power Points ”. The ’’Goldfish” card, provided by British Gas, has free spending of bonus points. The cooperation allows, for example, bonuses of banking services, groceries, telephone services, pharmacy products, clothing, bike and car accessories, insurance, etc.

Internet connection and other telecommunication services, design services and helicopter transport are other examples of non-energy -related offers. According to Andersson [4], to stay competitive, UK utilities concentrate on broadening their business and state that after adding gas distribution, telecom service is the second most interesting area.

An increased focus on certain customer groups can also be discerned. ’’Talking bills” or Braille for blind people, text phone and minicom for people with hearing problems and

23 Some of these services also include administrative IT- routines. 24 , British Gas and Scottish Power. 36 security functions, like password schemes at meterservice for the elderly, are some examples of theresidential sector.

As mentioned in section 3.4.1, new payment routines is another consequence of the deregulation. Offerings of pre-payment meters is one example. Projects of electronic payment routines are also merchandised, even though these systems are not yet launched. Comparisons of energy prices (gas and electricity) among different suppliers can be made at some sites, and discounts are sometimes offered. In Sweden, the power broker Svenska Kraftborsen AB25, provides price comparisons and electricity contracts to all customers. Also at thesite of Telge Energi price comparisons can be performed.

Shared and guaranteed savings is a concept of importance both in the UK and especially in the USA. Long-range contracts with guaranteed energy efficiency improvements and shared earnings, is an outcome of years of experience in govemmentally initiated DSM programs in the USA. The concept has now become a successful strategy to retain and win customers [23].

Real time pricing (RTP) to small customers is an area where Swedish ESCs so far are relatively modest 26. In the UK, it is an established practice that the customer gets a lower electricity price, while the supplier has the right to turn his supply off at times of increased electricity load. Price forecasts for every half hour the coming day are also distributed to companies to enable them to concentrate their energy consumption to periods of lower electricity prices. Thereby, the distributor can avoid costly power peaks.

Another form of RTP is practised in the USA [34]. There, one day ahead larger customers are provided with the electricity prices of the coming day. Thereby, the customer himself has the opportunity to control and concentrate his consumption to the economically most attractive times of the day. The function is also provided to residential customers by some suppliers. As transmission and distribution grids in the USA are often undersized, large savings can be made of both the utility and the customers, as the customer concentrates his consumption to low load periods.

As surveys have shown that the possibilities for energy savings are larger when automatic control is applied against the real time price compared to manual control [66], EPRI and ConEdison have commissioned Honeywell to develop a two-way communication system for load control and other information exchange. The two-way communication is by them seen as a strategic link for the power suppliers [27].

According to Augustsson [7], and the Swedish Technical Attaches [23], American VAS are mainly directed towards commercial and medium-sized industries. On the other hand, EPRI refers to an increased interest for residential customers, especially regarding IT-related services [45]. While the competition of the larger industries is rough with many initiated actors, the residential market is a market ’’waiting to be had”. Contrary to expensive company-specific solutions required of the larger industries, mass-produced

25 www.kraftborsen.se 26 The author has only found this service offered by Telge Energi. 37 solutions for residential customers will reduce the development and production cost: the more services sold, the better the pay off from the investments.

Many services that are emerging are communication-based. Internet access, billing and energy administration services on the Internet, electronic payment, real time pricing are the most common. But also smart house services occur. Functions included in the conception are energy management, security, telecom service and home services. Fields of applications are, according to Nyman [63], resource minimization, increased comfort, increased security and everyday organization. Through the development of smart house technology, IT also provides entirely new opportunities for controlling the process and following up energy utilization. It can be possible for the electricity suppliers to communicate with different equipment at the customers and that way both reduce the customers' energy costs and increase their comfort. The aspect of security is another important issue. The customer will know that the hotplate of the oven is turned off, etc. Experiments in providing such services have been performed in Walnut Creek, California (a joint venture with PG&E27, a cable operator, Telecommunications Inc. and Microsoft Corp.) Central and South West’s subsidiary, CSW Communications, intends to offer demand side management services through the network, including automated meter reading, customer messaging, in-home bill estimates and remote customer billing. It will offer the remaining network capacity to service providers, extending telephone, video, data and other information services.

Viewing the concept of smart house services, Nyman also claims that one system integrator is essential for the success of home automation. According to Leif Jansson, Sydkraft Elforsaljning, in-house services for the residences is a sector which will attract more companies than just the energy utilities. Telephone companies, Cable TV companies and energy utilities, that all have an established relation with the customers and a physical gateway to the residences, will compete to become the customers ’ service provider in this concept. Vattenfall, also has identified this market and claims it to become a billion dollar market annually [22]. According to Dagens Industri [76], the international telecom company Ericsson makes a similar forecast. The new smart residential gateways, which will enable the home services are predicted to be worth around 45 billion annually in five years. In the USA, some smart house services initiated by the energy utilities are already realized. The services offered are mainly remote control of certain equipment, burglary and fire alarm, climate control, Internet connection and electronic payment.

But American surveys also show that the customer ’s main interest is not energy-related, but entertainment-focused [27], and secondly ranked is security. By providing, for example, pre-pay-movies, Internet connections and/or alarm functions, the willingness of the customer to pay for the VAS increases. Thereby, they also contribute to the financing of the communication channel. The PC, the television set and the telephone are the most commonly used customer interfaces.

In summary, most energy companies have come to the conclusion that the only way to retain old customers and to attract new ones, without reducing their prices, is to broaden their business offerings. This can be done through cross-business strategic alliances;

27 Pacific Gas and Electric Corporation 38 thereby, theelectricity sales companies try to keep up the margins of the electricity price and also the size of their organizations. Influenced by the IT-era, many of the new services are based on two-way communication, either as a service itself offered to the customers or as an administrative tool to serve the customer and to coordinate the joint offer with the related alliance-partners.

3.4.2.1 Outlook That electricity has become accepted as being a commodity, with the typical price trend towards short-term marginal cost, forces the power industry to add new products and services to the power delivery. This tendency contribute to increased diversification of power industry. However, as all customers are not willing to contribute to increased electricity sales margins for the power industry, the created services themselves will probably in the long term also be sold separately and/or become the framework in new affiliated companies.

IT will have a central role in this development, partly to enable the virtual organizations required for the network and alliance-based value added services, and partly to be the key function thatenables the services themselves. As telecom is one of the new business areas focused on by many energy groups, IT will also acquire a value in itself and not only as an enabler for other services than communication.

Electricity sales companies might in the future, except for energy-related services, concentrate on the following IT service areas:

• integrator for smart house services • value network services, like credit card bonus systems • administrative Internet services • security • entertainment

Services for electricity consumers have no need to be energy-related. Based on customer enquiries, the chairman of EPRI (96) believes that communication will play an even larger role. In his vision, the utilities might even ’’throw in the electricity for free” at a customer purchase of, for example, a ”pay-for-view” movie, distributed by the energy utility or the telecom company. To that extent it might be a convergence of entertainment, telecommunications and energy.

Regarding energy-related services like DSM, the development of smart houses will also provide possibilities for direct equipment control against tariff or, when offered, the real time price. The services, which mainly will reduce the income of the grid owner (see 5.5.8), will be offered by the sales companies so that besides becoming the integrator, the ESC will also try to become the customer representative against the distributor.

Besides absolute VAS, in the concept of multi-utilities, communication services are identified as one of the most strategic development areas. Provided that power line telecommunication (PLT) on the low voltage grid has a breakthrough, which is very likely, the former distributors will get a natural role as the information intermediary. If the breakthrough comes late, within many energy groups, other subsidiaries will focus 39 on communication services through competing technical solutions. This is a tendency already noticeable in the USA. Customers have no interest in how their phone calls are transmitted, rather that they have a private line without any hearing problems and that the cost is reasonable. However, due to business, among the market actors, the distributor must be seen as the most natural operator of communication infrastructure; i. e. PLT, fibre-optic cables or other.

The electricity distributor or the subsidiary will here pick up the competition with the telecom and datacom industry. But utilized as a communication channel for services to households, it is even more interesting whose services the distributor will provide, for instance, services of the energy group, the electricity sales company, the telecom company, the datacom company, the Internet operator, or of the real estate companies, etc. In Sweden, for example, the telecommunication company, Telia, already offers energy supervision sendees to estate owners. Provided that home automation will increase as a result of technical progress, even IKEA in cooperation with an IT-partner is a conceivable competitor.

In the battle of the electricity market and the direction towards VAS, another combat is soon to come, for instance, who will become the residential customers ’ future partner, the integrator. Many industries have identified the residences, i. e. thehomes, as the next large market place. Ericsson has developed a residential gateway, a box that combines Internet solutions as well as equipment control. Philips have launched a TV-interface for Internet, and the linen-draper ’s industry, Electrolux and Philips, among others develop smart refrigerators and microwave ovens. To become the integrator, the telecom industry and the power business here possess the same strategic capital: a large customer base, a natural gateway to thecustomer and a long-standing customer relationship.

This highlights a critical issue regarding the home gateway, i. e. the organizations that have the required assets and competencies are not necessarily the energy utilities. Thereby, the need for strategic alliances and value-networks will become evident. Virtual organizations will be created where the utilities can supplement their capabilities through partnership. Leaders of American energy utilities forecast that the virtual organizations will become the new model for the power industry’s alliance-based VAS [71], made up of many interactive and independent entities.

With IT, the market will open up and a global market will be formed with no bounds to geographical conditions. Energy companies will, thereby, also expand geographically. The alliance-based virtual organizations will facilitate the expansion, while also physical networks in the form of local offices will become natural starting points in the marketing process.

The combination of Internet marketing and value-network-based services might also affect the marketing procedure of the energy groups. While Internet requires a relatively strong trademark, virtual network services, on the other hand, are based on the specific quality of the service rather than on a certain trademark. If the electricity sales companies manage to take the role of the coordinator, the diversification of the power business has succeeded. In the future, the electricity sales company might not launch its own competence and product, but rather the combination of the competencies involved in the network. 40 Even if the VAS strategy would fail, and the price focus only would become the most attractive to the customers, IT would still have considerable influence on the power industry. For example, administrative Internet services will soon become matters of course, instead of considered as VAS. Companies that choose to concentrate on a low price strategy will utilize IT for the realization. Intelligent search engines on the Internet or in the customers ’ own computers, will, for example, once a month "’check” the price levels of the different sales companies, perhaps the exchange market, and based on these, choose the least expensive alternative. This development has already been seen, for example, in the telecom business. Here, customers can choose the cheapest operator for each phone call only by applying a ’’box” between the telephone and the telephone jack. Thisservice is for example offered by Net-Net and Tele2.

Remote metering is another area where IT will also be further utilized independent of the VAS strategies. In the long term, customers will not accept preliminary charges. Actors that start offering exact invoices will force other market players to follow their example.

Finally, as Weiner states [83] ”In an industry in which power is increasingly sold as a commodity on open markets, the amount of information and computer technology to develop efficient markets will grow exponentially.” But it is also hardly possible now to predict all the new opportunities IT will provide in the future, just as five years ago, it was impossible to anticipate the development of the Internet or 15 years ago, the penetration of mobile communication.

3.5 Examples of Swedish Communication-related Projects driven by Energy Companies The sales companies of the main Swedish electricity suppliers, Vattenfall, Sydkraft and Stockholm Energy, are all involved in IT-related projects aiming to provide the customers with IT-based value added services. Below follow some examples of these activities.

3.5.1 Cooperation with Nor.Web To provide Internet access for residential customers, Stockholm Energi, Sydkraft and Vattenfall separately have made field tests of the Nor.Web patented technical solution for broadband communication (called DPL, Digital Power Line) on the low voltage grid. The British-Canadian Nor.Web is the outcome of a joint-venture between the Canadian Nortel (Northern Telecom) and the British multi-utility United Utilities. Satisfying test results enticed the competing Swedish energy groups, Stockholm Energi and Sydkraft, in alliance with Tele2, into together approaching Nor.Web for an agreement regarding technical utilization.

Nor.Web has achieved communication up to 1 Mbit/s on the Swedish low voltage grid in a range of 300 meters. Since there are problems in passing the transformer of the net station, the signal is sent from there by other means, for instance via copper signal cables or optical fibres. According to the test results, there are still other technical

41 limitations in the concept, but so far Nor.Web seems to be one of the most successful companies in this field.

The possible profitability of DPL technology is much dependent on the pricing of competing broadband technologies. Thus, by using existing infrastructure, i.e. the low voltage grid, the chances for a commercial breakthrough for DPL are enhanced.

In England around 100 households, customers of United Utilities in Manchester, are already supplied with the Nor.Web solution for Internet access. Also 200 customers of Energie Baden-Wurttemberg AG (ENBW) in Karlsruhe, Germany, are testing the Internet access via power lines.

With this solution, other communication services can be introduced as well. But according to market surveys, Internet extension is one of the few VAS that residential customers are willing to pay for.

3.5.2 Gotcom 28 The governmentally owned Vattenfall also performs field tests of other IT applications. On Gotland, an island in the Baltic see, Vattenfall has introduced IT-based applications to around 80 customers. Gradually all the 36 000 households on Gotland will get the opportunity to try out the new Vattenfall services. The investment in the project is around 140 Mkr and extends over three years. The project has two separate aims: partly to create value added services through customer communication and partly to make Vattenfall a more effective grid operator. The basic idea is communication on the low voltage grid and lon-works technology inside the buildings. According to Vattenfall, the project shall be seen as the embryo to the distribution company of the21st century.

The concept, to begin with, is built around the electricity suppliers’ only physical connection to their customers, namely the electricity meter, but probably also around the requirement of hourly metering of all residential customers on the point of changing supplier. Thereby, the electricity meter has been given ist central role. The target cost for the VAS is 2000 SEK/customer which can be justified by the metering function, as Vattenfall has promised all residential customers hourly metering. Today a customer has to pay around 2500 kr for a meter. In addition, Vattenfall has chosen a ’’smart” electricity meter. Besides the sum meter function, it will be able to differentiate among the different equipment within the house. For instance, it directly alarms if something seems to be out of order. Future electricity bills might, thereby, instead of incomprehensible tariffs, present the cost for warm coffee, cooking, microwaved food, electrical heating, lighting, etc.

Packaged VAS that will be introduced are Electric Heating Specification Packages, Tariff Specification and Administrative Routines, DSM Concepts and Security Packages. In detail, the following services are included: Web Interface, Energy statistics

28 Vattenfall has published very few articles concerning the Gotcom project and is unwilling to answer questions regarding the case. This probably implies that the project is of large strategic value and a main cornerstone in their "added value ” strategy. Most of the information above is taken from a seminar where Gotcom was presented, at the DA/DSM conference in London, October 1998.

42 and Consumption Data on the Internet, Environmental feedback and fast feedback to customers with changed consumption patterns, Detailed nonpreliminary tariff specification, Simple Invoicing, Administrative Payment Options, Load Management and DSM, for residences with mainly the disconnection of boilers at overload, Equipment and Climate Control through smart house technology, Alarm Functions for establishments, movement-detector-based burglary alarm and Automatic Earth Fault Contact Breaker.

Another area is to perform secure transactions over the Internet. Such applications are said to be also tested within Gotcom.

Smart refrigerators are also included in the vision, as the customer interface. With the ’’Kajsa Varg - function ” the refrigerator will automatically respond to customer needs by grocery ordering or by creating recipes based on its contents. It might also be possible to phone the fridge from the grocery store to find out what’s missing.

As the services are introduced, gradually the electricity meter will become only a small part of the many services Vattenfall offers. Thereby, Vattenfall has started the process of becoming a service provider, namely the integrator, for their customers, instead of only being the supplier of a commodity. Even more interesting is their future intention to offer these services to non-electricity customers.

Regarding more efficient grid operation; extension of traditional SCADA - DMS (Distribution management systems) the following systems is tested.

Further, the project also includes testing of Business Process Reengineering (BPR).

3.5.3 ISES29 In Ronneby, several market players together have initiated the ISES project. Among them, Sydkraft, Preussen Electra, Electricite de France, ABB, IBM and IT can be recognized. The local distribution utility, Ronneby Energi, is also closely involved. ISES aims to provide research results within ’’IT and Energy” to be utilized by the sponsors. The project covers a wide area of potential IT utilization within the power industry. In the beginning, it consisted of nine different ’’Sub-projects ” regarding

• Cost Optimization of Energy Systems - profitable measures30 • Communication on Low Voltage Grid • Distributed Load Control • Distributed Autonomous Decision Islands • Computer Control Technology for Ronneby Energy • Databases and Structures of Documents • HMI Interfaces/Information Kiosks • SecurityIssues concerning the Information Society • New Business Strategies

M Information Society Energy System 30 This thesis is to be seen as the final report from this subproject. 43

% • Decision Style Approach to Information Processing and Interactive Marketing • Virtual Organizations

(For more information, see 5.3.)

3.5.4 ID AM In 1994, Sydkraft started the power line communication project IDA (Intelligent Distribution Automation). The aim was, in the beginning, mainly to obtain sufficient knowledge for strategic decision-making regarding long-term IT system extension. Two-way communication between the supplier and the customers was realized on the low voltage grid, and local ’’intelligence ” was placed in the distribution grid for more efficient distribution operation. To start with, focus was laid upon load management at distribution limitations and remote electricity metering.

In 1995, cooperation with IBM was entered concerning hardware solutions, and IDA became renamed to ID AM (Management). When the market became deregulated, customer-orientated services were added to the concept. Products for property owners containing climate control, temperature measurements, automatic fault indications, control of oil consumption against electricity prices, etc. were developed.

ID AM is limited to relatively slow data transmission because of the small band communication technique, and thereby, mainly on and off signals can be transferred. Further, due to high component costs 31, one actor alone cannot afford to introduce the technique. Cooperation among the different actors, the electricity sales company, the network operator and perhaps even more players is essential for large-scale realization. For how long this will remain the truth is hard to predict, as component and communication costs continuously decrease.

3.5.5 The Sydkraft’ ’Spjutspetsprojekt ” In order to find and evaluate new business opportunities as the margins of the electricity sales are reduced, Sydkraft initiated in 1998 their ’’Spjutspetsprojekt ”. Totally focused on VAS, the project utilizes strategic alliances and IT applications to create a technical platform which will enable the new services. Alarm, Internet access, e-mail, remote control of certain equipment, climate control, secure payment routines, smart cards, entertainment, like downloading music, videos and computer programs and every imaginable service connected to the brisk trade of Internet, might become a new offer. The television set is chosen as the interface.

In the creation of the ’’electronic home”, the technical advances that will enable the development are taken for granted, such as,

• Fast communication channels to the building • Fast, flexible LAN inside the building and corresponding WAN in the absolute nearness

Prices of 1997 44 • Intelligent, more specialized and network connected applications within the homes • Convergence - Datacom, telecom, audio, video, process communication, etc. will utilize the same media and protocols

This means that all equipment within the homes will be able to communicate with all other equipment within the homes and within the world. According to some of the alliance partners (Ericsson, Nokia), the technical conditions will be realized within one to three years.

Sydkraft intends to become the coordinator of all the new alliance based services that will be provided in the households, an integrator that guarantees reliability and security for the customer.

3.6 Communication The telephone lines is probably the most frequently utilized communication infrastructure within the power industry followed by radio and radio link. Also PLT is utilized but is still limited by thetransmission stations either to the low or to the medium voltage grid. Fibre-optics seems to be the coming technique as it allows secure information, even at large data transport. Operations in the grid, like control of remote switchers, are often enabled by radio due to location.

4. METHODOLOGY

4.1 Performance Interviews, literature surveys, comparative studies, internal company evaluations, energy system optimizations and industrial simulations of real cases constitute the base of this thesis. The following activities have been performed in a period from 1996 tol999.

Literature survey/interviews Consequences of deregulation Determination of electricity prices Strategies in electricity sales

Interviews Current utilization of IT within power system balancing, metering, distribution supervision and electricity sales

Literature survey Future potential IT utilization within the areas above

Comparative study VAS launched at British and Swedish energy companies homepages on the Internet

45 Sydkraft internal evaluation IT-related VAS, economical benefits

Energy system analysis Application of MODEST on the Ronneby Energy System, long-term analyses with traditional tariffs

Application of MODEST on the Ronneby Energy System, diurnal analyses of load management against the spot price versus traditional tariff

Industrial simulations Application of INDSIM on two major industries in Ronneby, annual analyses of DSM measures against modem electricity contracts (after the deregulation) versus traditional tariff

4.2 Modelling For the calculations, the energy system optimization model MODEST has been applied on the municipal energy system of Ronneby, and the industrial simulation model INDSIM has been used to mirror the benefits from DSM measures in two major industries in Ronneby. For descriptions of the models, see the enclosed Paper I (MODEST32), and Paper II (INDSIM33). The overall aim with the analyses was to evaluate eventual benefits from IT-based load management. Three different types of electricity prices have served as boundary conditions: firstly, monopoly tariffs (1996- 1997), secondly, sales and distribution contracts of the competitive market (1997) and thirdly, refined prices of the spot market (1997); thereby, also consequences from real time load control against the power exchange have been examined. IT-based end-use measures are simulated in all cases.

In the industrial analyses, the customer ( C ) is put into focus. The simulations show the customers benefit from load control corresponding to the different market conditions of monopoly, competitiveness and trade market. The impact on the supplier, meaning the distributor (D) and, when called for, the electricity sales company (S), from these measures is also mirrored. Finally, eventual benefits, i.e. united system earnings, from cooperation between the supplier and the industry are estimated.

By applying MODEST on the entire energy system of Ronneby with monopoly tariffs as input, the focus of this study is put on the costs of the distributor/sales company. Also here, the IT-based load control is seen as a cost effective cooperation form between him and his customers. It is examined how the control affects the utility in the optimal energy system, and the consequences upon the regional power producer (P) are discussed.

32 Further information about MODEST can also be obtained in the licentiate thesis ’’Energy Systems Optimisations Applied to Local Swedish Utilities”, Henning D., se references 33 Further information about INDSIM can also be obtained in the licentiate thesis ’’Cost-Effective Incentives for Local Electric Utilities and Industries in Cooperation - Modelling of Technical Measures”, Andersson M., se references 46 Finally, the diurnal optimization of load management against power exchange market spot prices show eventual differences in load control utilization corresponding to different price structures.

5. CASE STUDY RONNEBY The Ronneby case study is to be seen as an effort to bring down the large upheavals of the deregulation and the impact of the IT-era to a smaller, local utility perspective. How the deregulation has influenced the local utility and if information technology can improve the optimization of the municipal energy system are questions to be answered. The results referred to are mainlybased on the enclosed papers.

5.1 The energy system of Ronneby The city of Ronneby is located in thesoutheast of Sweden, 550 km south of Stockholm. When the energy system analyses were first performed, the local public utility,Ronneby Energi AB, (REAB), distributed around 250 GWh electricity and 33 GWh heat in 1996. The maximum demand was 54 MW power and 10 MW heat.

In 1996, the electricity was purchased from Sydkraft due to long-range agreements, made up before the deregulation. Today, different suppliers compete for the industrial customers, and the Norwegian oil and energy group, namely Norsk Hydro, has taken over the supply to the residential customers.

The picture 5.1 shows the energy system of Ronneby in 1996, including the options analysed and described in Paper I.

Briefly, these first analyses concluded that, apart from different end-use measures, building of a bio-fuelled combined heat and power plant (CHP), would considerably reduce the cost for supplying Ronneby with heat and power. With the investment costs included, the cost reduction due to a new CHP was calculated as 30 MSEK in five years.

47 ENERGY SYSTEM OF RONNEBY 1996

existing equipment Eol = Light Fuel Oil imi potential investments B = Boiler

Figure 5.1 Ronneby Energy System 1996, with the potential measures analyzed in Paper I

48 Since the investment costs for small-scale CHP seldom enable cost-effective solutions, these solutions should only be applied under very special conditions. For the analyses, the heat demand was increased to 60 GWh, which also corresponded the then approximated future demand for district heating. The picture 5.2 shows the relation between investment cost and CHP production capacity. The graph is based on data from realized production plants.

35000

30000

25000 __ ♦- 20000

15000

Electric Power capacity (MW)

Figure5.2 The relation between CHP investment cost and electricity production capacityu

Several analyses have been performed of potential measures in Ronneby. Due to the interesting results from the first simulations referred to in Paper I, on a consulting basis, further analyses were performed especially concentrating on the potential benefit from an investment in a hot water plant, compared to the benefit from the building of a combined heat and power plant. Theoptions illustrated in figure 5.3 were examined.

The technical consulting company SYCON Energikonsult, where the author is employed, is the company that performed the analyses*35 [10]. Based on MODEST, SYCON has developed the four-steps-concept ’’The Energy Steps”, which has as its objective to give technical support to the customer, all the way through the realization of a new production unit, i. e. from the abstract idea to the actual building. Energy system analyses with MODEST constitute the first step, 1), of ’’The Energy Steps”. The MODEST analyses contribute with the base on which decisions can be taken regarding further, more detailed, technical and economical analyses of the most attractive solution or solutions. The other steps of the concept are: 2) A Feasibility study, which result in a proposal for the eventual further development of the energy system, 3) Planning. This is a detailed planning for the realization of the chosen solution, where the technical and financial consequences are evaluated. 4) Purchase. In this phase the engineering and

31 Source: NUTEK, 1997 35 The author performed the analyses together with colleagues at SYCON. 49 contracting work for the new plant is performed.

Sydkraft power

Industry

Electricity Services, etc. distribution

Residences

Load Street lights |

Total Heat load (72 GWh)

Potential equipment Existing equipment

Figure5.3 Further analyses of the Ronneby Energy System

The MODEST analyses recommended the building of a 12 MW Hot Water Plant, but in the following evaluations, also the alternatives of figure 5.4 were examined. For all cases, it was assumed that one LPG-fuelled boiler of 8 MW and three oil-based boilers of 3x8 MW would cover the required peak demand.

50 Fuel Plant Efficiency Alfa value 36 Investment Fuel price* Capacity

1%1 [kSEK/kW] [SEK/MWh] [MW] bio CHP 1 87 0.35 20-25 120 4.2/12

bio CHP 2 102 0.28 20-25 120 3.4/12

bio HWP 102 - 3.5 120 12

light oil HWP 90 - 1.0 330 3x8

LPG HWP 90 - 1.0 280 8

* operation and maintenance costs and taxes included

Figure 5.4 Table of potential investments

Ronneby was here analyzed based on the estimated total heat demand of 67 GWh. Given distribution losses of approximately 7%, the required heat production was estimated to 72 GWh. The assumptions were based on the answers to questionnaires from Ronneby inhabitants.

With a utilization time of 2250 hours, the power peak of 32.3 MW was given. But, since reserve boilers, are to be utilized to cover the peak demand, to avoid a too large investment in heat base production, the peak demand in the energy system analyses was approximated to 75% of the maximum peak, i. e. 24 MW.

The analyses were further based on an electricity demand of 250 GWh electric energy and 54 MW power peak. The discount rate was 6% and the analysis period five years. The depreciation for the investments was 20 years.

Given the conditions of the latter energy system analyses, these calculations concluded that the building of a bio-fuelled hot water plant (HWP) of 12 MW was the most interesting alternative, provided that there was less than 5%/year electricity price increase. At no price increase, the energy system cost of Ronneby would be reduced by almost 60 MSEK in five years. The large reduction was mostly due to high oil prices for residential customers.

Figure 5.5 shows the energy system costs due to different investment alternatives, and the dependence on an eventual electricity price increase. At 5%/year price increase, a bio-fuelled CHP with the investment cost of 20 kSEK/MW, would be as profitable as the HWP.

36 Electricity to heat ratio. 51 However, the following analyses of Energistegen showed that due to the stepwise expansion of the district heating distribution network, starting with a 6 MW bio-fuelled boiler would economically be the most beneficial.

In December 1998, a new bio-fuelled heating plant of 6 MW, plus 3 MW from flue gas condensing, was inaugurated. As mainly the heat from residential oil-based boilers was replaced, the electricity demand in Ronneby has stayed about the same. The heat demand and peak are in the long term forecasted to increase to approximately 75 GWh and 20MW, respectively, per year, and REAB will, therefore, gradually continue with the realization of the 12 MW HWP proposed in the latter energy system analyses.

Energy System Cost [MSEK]

-. --J* . —HW P bio —a - Existing system C H P 1 —.— C H P 2 450

430 tjf-

tquai system cost

1 1

1 350 4=

Electricity price increase |% ]

!

Figure 5.5 Energy system cost due to different investment alternatives and potential electricity price increase

52 5.2 Impact of deregulation As Ronneby is located in the south region of Sweden, Sydkraft was their natural supplier of electricity in the monopoly era. Early, already in 1991, discussions regarding regional cooperation between the municipal energy utilities of Ronneby, and Karlshamn, were started. The aim was to obtain coordination of earnings in terms of lower operation costs and joint purchases at the controlled electricity market. But, the driving forces of the owners were small and the issue was left. This was perhaps also a consequence of the regional joint failure in trying to purchase thenearby located energy business of Alfa Laval. The business was later purchased by the Olofstrom municipality.

First, in January 1996, at the same time as the introduction of the deregulation, the regional electricity sales company Blekinge Energi was formed from the market divisions of Ronneby Energi AB, Olofstrom Energi AB and Bromdlla Energi AB. The object was to win large purchase advantages through increased electricity volume. The following spring, Blekinge Energi entered into negotiations with KREAB, a local electricity sales utility located on the southwest coast of Sweden, regarding potential cooperation. At the time, KREAB directed their efforts towards power trade and towards spot price -related power agreements in cooperation with Sydkraft. The discussions came to no result, and during the summer, the municipal owners of Blekinge Energi decided to sell the company. This was motivated by the delicate question of whether speculative electricity trade is a municipal concern. The conclusion drawn by the municipalities was evident as the Norwegian oil combine group, Norsk Hydro surprisingly bought 80% of the utility later in the autumn. The given objective was to establish their operations in the electricity market of . Later, in the beginning of 1998, the final shareswere purchased.

In addition to the changed proprietor structure, also many customers have changed supplier due to the deregulation. For example, Cetetherm, which is part of the Atlas Copco group, has through a combine group agreement from 1 January 1999 obtained Vattenfall as their supplier. Also, Volvo in Olofstrom is through similar arrangements supplied by Vattenfall. The large plastic flooring manufacturer, Tarkett, has made an agreement with Graninge. Of the largest industries, only Richer Folie has chosen to stay with the former Blekinge Energi , i. e. Norsk Hydro. Also smaller, non-industrial, customers, have changed supplier. For example, the Ronneby Brunn hotel, has chosen Birka37, for their electricity deliveries. Regarding residential customers, roughly 200 of around 4000 house owners, with electric heating, have changed supplier, and even more movements are expected as the meter requirement is taken away.

Regarding the remaining distribution utility, Ronneby Energi AB, the company has expanded their operations. Now the business, except for electricity distribution, also includes electricity purchases for the and district heating. Further a merger is to be expected between REAB and other municipal businesses. Later in the spring of 1999, the authorities of water supply and refuse collection and disposal will be integrated with REAB. The name of the new company is yet not decided.

Another new increasing business of REAB is placing or renting infrastructure for communication services. REAB gets many requests, mainly from tenant-owner

” Birka is owned by the Stockholm municipality and the Finnish Imatran Voima (IVO) 53 societies, regarding the placing of fibre-optic cables. Therefore, they have decided always to place extra cable pipes at the same time as placing pipes for district heating. Further, the renting of this communication infrastructure to customers is becoming a remunerative enterprise.

Looking at communication services, READ is also active within the ”IT and Energy”- related project ISES, see section 5.3. ISES aims to obtain research results applicable within the power business. Information technology is here seen as one of the major factors thatwill influence the future power industry. The object of REAB’s participation in the project is to introduce new valuable services to their customers. Consequently, these will contribute to even further expansion of REAB’s business concept. A first natural step is to start with meter data collection and electricity consumption statistics. This is already performed in Patorp, a residential area, within Ronneby. Latter services are equipment control and security offers, like fire alarm, burglary alarm and leakage alarm. ISES also contains projects of energy system analyses. For example, REAB’s investment in district heating is partly a realized result from system optimization performed within this project. Below follows a summary of research activities within ISES.

5.3 Energy-related IT-projects - ISES With the distribution utility Ronneby Energi as the experimental arena, the ISES project started in 1994. Several market players, Swedish and foreign, sponsored the project 38 . The aim of the larger energy groups was partly to study the impact of the deregulation and partly to examine the potential for and benefits from increased IT-integration in the power industry.

ISES covers a broad IT perspective, from marketing strategies to advanced technical issues. Defined as interesting research areas, within the field of distribution, were distributed load control, autonomous decision island and computer control technology. Regarding communication infrastructure, the potential of the low-voltage grid was highlighted and concerning the electricity sales procedure, research of new business strategies, decision styles and virtual organizations got started. Other topics have been viewed within ISES as well (see 3.5.3 ), but will not be further discussed here.

Apart from the energy utilities as important actors regarding the realization of the information society, two main reasons are identified within ISES for the engagement of these actors in information technology. Internal efficiency improvements through distributed automation and external customer-oriented services are key issues for whether distribution utilities or electricity sales companies engage in IT expansion. Both areas are considered to contribute to increased yields.

Within ISES, it is expected that tomorrow ’s energy system will consist of a large number of communication devices operating close to the customers. Load control will be included in the service spectra offered, and thereby, increased system optimization and increased utilization of present distribution assets enabled. Within the ISES

38 ABB, IBM, EOF, PEAG, Sydkraft, IT Blekinge and Ronneby Municipality. 54 subproject of distributed load control, it is shown that it is possible to perform automated sophisticated distributed load control, even with very large numbers of loads [6], [86]. Thus, a step is taken towards real time load control. Advanced software agent technology is utilized and the mechanism for the control is based on a enquiry/offer/decision method. The object is to reduce the desired demand withthe least cost and inconvenience to the customers. Theoretical models have been derived and simulated, and some initial field tests have been performed within this subproject of ISES.

As for distributed load control, also regarding implementation of VAS through autonomous decision islands, analyses with theoretical models and initial field test verifications have been performed in another closely-related ISES subproject. This research has shown that it is possible to implement VAS, like improved energy efficiency enabled by PLT between agents and electrical devices of a building [40]. Simulations indicate energy savings up to 40%.

Besides energy-related VAS only, other communication-based services can be introduced by the distributor or the electricity sales company as well. New business strategies and how to achieve increased precision in mass marketing are therefore other topics within ISES. Research has been carried out regarding a decision style approach in these matters. For example, how to build an interactive utility information service on the Internet has been investigated. Flexible complexity levels and interactivity depending on decision style are included to create learning systems for the service provider. The overall objective is to reach higher precision in the service development and the marketing process.

In a closely-related project, the decision style model also contributes to increased understanding of information dissemination in the cooperation among organizations. Another consequence of the deregulation is, as discussed in section 3.4.2.1, the emergence of virtual organizations. Viewed from the perspective of decision styles, the ISES subproject 4 aims at enhanced interpersonal communication through mutual understanding of how people process information. Other issues to be investigated are the role of IT in virtual organizations and how these organizations learn across corporate boundaries.

5.4 Enabling IT infrastructure The dialogue with the customers and the discussed value added services referred in section 5.3 are all IT-based. That these new communication services will be introduced to residential and industrial customers within a few years and that the customer dialogue will change and increase is very likely both due to technical progress and the free market. A more urgent question is who will be the provider of these services and what kind of technique will be utilized. In section 3.4.2.1, this topic is referred to along with the different market players’ interests in the residential market. Provided that the electricity distributors take the lead, power line telecommunication (PLT) is likely to playan important role in the realization.

According to Akkermans, Healy and Ottosson [5], PLT also has one unique advantage over other communication techniques: it constitutes an already existing infrastructure to 55 billions of residential customers and businesses. Further, it also connects equipment within the homes.

Utility I Gnd_companyP' ^Sale company 3 Utility communication 1 networks Info- Low voltage grid kiosks Lan radio IR-technology

Municipal communication networks 3 3 Internet - other networks

Figure 5.6 Potential future hybrid infrastructure for communication of the information society

The future communication infrastructure of the information society will most likely consist of a combination of techniques. In Ronneby, as mentioned, the utility expands the network of fibre-optic cables for customer communication. PLT is also partly utilized for metering purposes and collection of meter values. Cable TV, telephone and radio are established communication techniques. It can further be assumed that the existing infrastructure will be utilized to its furthest extent as long as the existing communication technique enables the required information transmission capacity. The picture 5.6 illustrates a potential future infrastructure to enable the information society39 .

5.5 IT - a condition for optimized energy systems

5.5.1 Postulations Optimized energy systems require real time response on changes of boundary limits. For minimization of the energy system cost, in the optimized energy system, these changes need to be converted into price signals. It is further assumed that all actors directly respond to a given signal, even though this might not always be the case in reality.

Based on a cost perspective, optimization of energy systems is considered to be equal to maximum economic utilization of existing resources weighed against potential new

39 Source EnerSearch AB, SE 205 09 MalmO, Sweden 56 assets or measures required to supply a specific demand. This means that all equipment, generating units, distribution assets, or potential controllable load, etc., within the system limits, needs to be considered to obtain the minimum energy system cost.

The discussion below concerns worldwide, national, regional or local energy systems. The common factor is that electric load is seen as a variable unit (partly controllable) instead of only a constantly growing demand. To obtain an optimized energy system, in the text, eventual legal restrictions are not taken into consideration.

5.5.2 Load control - essential for minimization of energy system cost The results from Paper I, show that the energy system cost decreases considerably when the load control is introduced; thus, load control contributes to improved utilization of energy system assets, i.e. ’’more ” optimized energy systems.

The figure 5.7 shows how load management in Ronneby would be utilized according to thetime division of MODEST.

MW i Winter peak da\ Load Management

Winter —- 22-06 Apr. Sep, Okt week days 22-06

Spring

Summer Apr, Sep, Okt week days 06-22

New load curve Former curve

Hours

Figure 5.7 Power Demand in Ronneby with Activated Load Control

The reduction of the energy system cost is a consequence of the time dependent supply - demand functions of electricity generation, S=f(t), and electricity consumption, D=fft).

57 Economic boundary conditions settle whether it is beneficial or not to control a certain unit.

kr/MYVh

MWh

Figure 5.8 Supply demand graphof load control at the time tr

In a supply-demand graph, load control for each time period can be visualized as a move of the demand curves to the left. The constraints for optimized load control are that the energy transpositions of the concerned period are equal

X3* - X[- = Xy- - X]-' and that the price reduction (y3- - yr) is larger than the price increase (y3- - y,-).

y3--yi-> yr- yr- t2-

kr/MWh

MWh

Figure 5.9 Supply demand graph of load control at the time t2-

58 When only industrial energy-related equipment, like electric boilers, are controlled, the energy system cost decreases with approximately 4 MSEK in 5 years. In Paper I, it is shown that when residential load, such as, hot water heaters and dual-fuelled boilers, also are included, thecost is reduced by more than 12 MSEK in 5 years.

Thus, the key to an optimal local energy system is the ability to control load. Further, to obtain an optimized energy system operating on the limits of its boundary conditions, all potentially controllable equipment must be accessible for the comparison of the system benefit from load control versus the SRMC of electricity production. This, for example, includes both residential hot water heaters and process-related equipment within industries. Viewed as single units, the system contribution from load control of one hot water heater is more or less negligible. In contrast, due to their large number, for the system optimization, residential equipment become as important as any controllable power demanding equipment within an industry.

In a monopoly market, the utility can optimize its income by managing exact amount of load due to fixed boundary conditions of costs and income of power and electricity purchase and sales. For example, at sudden power peaks, the utility can avoid reaching its power purchase roof by activating load management of concerned customer equipment. To attain the optimal solution for the distributor - he has to control the exact amount of load at theexact moment, which probably only is achievable by managing the equipment of several customers. It is also likely that the energy -related equipment at different industries, such as, electrical boilers, heat pumps, etc., cannot provide the exact amount of load required from the distributor at the exact moment. Further, operating on the capacity limits of an energy system, theoretically many small steps, instead of one large one, enable improved flexibility in system operation and, thereby, optimized energy systems.

5.5.3 Customer incentives for load control Some general postulations can be made regarding the market actors' incentives for load control, independent of if the market is characterized by monopoly or by competition:

1. The electricity supplier always wants to earn as much money as possible on his electricity deliveries. This means that he, depending on his purchase and sales agreements, always wants to deliver the exact amount of electricity and power that optimizes his own income.

To exemplify: The distributor wants to deliver as much electricity and power to his customers as possible, as long as he has a positive and/or increasing contribution margin and does not reach his own power and electricity purchase limitations.

59 2. The industrial customer always wants to earn as much money as possible in his business. He is, therefore, willing to introduce all legally correct and cost-efficient measures thathelp himto maximize his income.

To exemplify: As the industrial customer wants to optimize the income from his business, he is also interested in reducing his energy costs. But he is only willing to introduce technical measures that ensure the optimization of his total earnings, i. e., that his main income source, namely the production, is not disturbed.

Thus, by definition, the customer is willing to manage all equipment as long as the measures contribute to reduce his energy costs and do not reduce his comfort, earnings or object of business.

In Papers I and III, the load is divided into three categories:

• residencies • industries • service and miscellaneous

Outcomes from the papers show that the economic benefit from load control varies among the categories and depending on the amount of load controlled. The customer benefit in the residential sector is rather small based on current electricity price levels. By control of the hot water heater and the dual-fuelled boiler against the tariff the customer earns around 70 SEK in one year. Since equipment enabling the load control according to price or power fee is currently expensive, and few incentives are given in existing electricity price structures, residential customers are unwilling to pay for extra equipment to enable this control.

In contrast Paper II shows that the customer benefit from industrial load control is high and that it sometimes even has increased due to new power agreements of the free market. It is also shown that within an energy system, where the system limits correspond to those of the local distributor, cooperation between him and his customers can lead to shared savings in terms of commonly reduced power-related charges.

According to general science, load control and energy savings are of large interest for most industries, as long as the measures do not disturb the main production process. Two main load categories of controllable equipment can be classified regarding the industrial customers:

• Energy-related equipment (for heating, etc.)

• Process-related equipment

The first, like heat pumps and electric boilers, is often the subject of load control arrangements based on agreements of the electricity delivery contracts. The equipment concerned is of little importance for the customers ’ main process. Process-related

60 equipment, on the other hand, is of large importance for the customers ’ business core. Eventual interruption here might cause major yield drops, for instance, for the manufacturing industry. Experiences from projects driven by the Division of Energy Systems at Linkoping University reveal that the industrial interest in these matters mainly arises first when the potential cost savings are calculated, optimized or simulated. But as energy consumption mostly only stands for a smaller part of industrial costs related to the yield, often around 3-5% [80], focus is instead rather laid on the main income source, the production process. Interruption here might become very costly. Therefore, the industry is only willing to manage process-related equipment, if he is ensured that the measures do not disturb his production process. Thus, control of this equipment is reluctantly left to an external part. If process-related equipment should be taken into account for remote load control to obtain optimized energy system conditions, the activation would, therefore, have to be initiated by the customers themselves depending on their production plans and obligations. As the customer himself makes the beneficial judgement, i. e. in real time optimizes his earnings based on thecosts for an eventual production break versus the potential decrease of energy and power costs, his cost savings are weighed against his eventual loss of income.

5.5.4 IT requirement Out of practical reasons, specific equipment of residential load cannot be manually controlled at all times during the day and year. Further, the amount of potentially controllable equipment, for instance, for climate regulation within building complexes, requires continuous supervision and simultaneous control, which also leads to digital information transport. At a deregulated market, characterized by its right of options, price signals or former agreement must constitute the base for this control.

When power price exceeds a certain level, the equipment is, by the customer ’s choice, disconnected, below this level the customer prefers to utilize his equipment. Thus, the price of electric energy and power, together with the customer ’s economic dependence on his equipment, will determine how theenergy system can be optimized.

Therefore, due to the essential amount of potentially controllable equipment either manually unreachable, like hot water heaters, or unavailable without option of timely price comparison, like process-related equipment, information technology is required for optimization of energy systems.

Pointing at the definition of optimized energy system as based on maximized economizing of existing resources, digital information transmission is required as ’’bids ” are thetypical information transferred. Change to analogue technology can here only be seen as an extra cost, since the digital information would have to be encoded at the sender and decoded by the receiver into representing physical quantities, like frequency or voltage.

5.5.5 Tariff construction versus system optimization It is important though to remember that given the existing conditions of the deregulated electricity market, where only the electricity sales companies have been exposed to competition, there are few incentives for the still monopoly distributor to reduce his 61 customers ’ energy and power costs. Doing that, he also reduces his own income. The main reason for distributors to introduce load control is, therefore, when there are capacity limitations, like distribution bottlenecks, or to minimize the impact of potential actions taken by the electricity sales company (see chapter 2 and section 5.5.8).

In an optimized energy system, it is shown that IT is required for utilization of load management to compensate timely variations of, for example, SRMC of electricity production or at distribution limitations. Viewing the spot price variations and the regulate market of the power exchange, it becomes evident that the time tariff does not mirror the energy system dynamics. Therefore, load management against time tariff might lead to system suboptimizations. This is shown in Paper II and also by Andersson [2], [3].

To enable real time optimization of energy systems, timely correct incentives for DSM measures must also be transferred to the customers. This means that the pricing of electricity and power to the customer must correspond to the timely variations of production capacity and distribution limitations. Thus, real time pricing is a condition for optimization of energy systems.

To enable real time energy system optimization in a free electricity market, power and energy prices should each be based on the real time relation between supply and demand. Power prices should be a timely function of maximum distribution capacity and current load. Similarly, energy prices40 should correspond to the timely price of activated production resources within thesystem.

5.5.6 The optimal energy system Viewing the power exchange, it is noticeable that the spot prices follow the demand with a slight delay. Thus, energy systems are dynamic, and electricity prices are determined by the price cross of supply and demand. However, there are activities within society that need to continue whatever the boundary conditions of energy systems. For example, respirators at hospitals are not disconnected at almost any electricity price. Mathematically, this can be illustrated by giving these activities an infinite cost for potential disconnection. Thus, even if most equipment within a system can be rescheduled, the energy system also consists of a ’’lowest ” electricity demand level that can be reduced neither by improved efficiency nor by conversion to other energy carriers.

Though, to reduce the energy system cost to the greatest extent, all controllable load should be rescheduled to low price periods. In a perfect market, the energy system dynamics will compensate this move though, with higher prices until both the real time price and theload curves level out.

Provided a perfect market and that most energy consuming equipment and generating units can be balanced against each other, the optimal energy system would obtain a price and load curve similar to the curve LCn in figure 5.10. Thus, a consequence of total optimization is the levelling out of both production and demand profiles.

40 Real time energy prices are similar to power prices based on the available production capacity 62 (kr/MWh)

Regulate Hydro Power

Nuclear Power

Hydro Power

hours

Figure 5.10 Illustration of the duration graph of an optimized energy system

In Paper III, the consequences from only spot market-related load control within a small customer group are viewed. Here, the load and price curves do not level out, which is shown in figure 5.11.

63 Spot price related load management 60000

50000

40000

5 30000

20000 -

10000

- - - Original load curve —Manipulated load curve ■

Figure 5.11 Illustration of optimal load control for the customers against the spot price

In such a case, without any active power charges, the impact upon the distributor might become considerable.

It is further shown that current power-related charges determine how and when load is to be controlled within a system. Levels of, for example 200 SEK/kW in the distribution contracts are high also compared to the electricity price levels of the regulate exchange. Here, the price often varies around 0.25 SEK/kW, see 5.5.8.

The price construction of the distribution contracts, where power-related charges first of all correspond to the daytime peak demand during the winter season, results in few incentives for continuous load control in other periods. Therefore, when the power charge is charged, mainly load rescheduling between day and night remain in those cases where time tariff is applied.

5.5.7 One- or two-way communication It is shown that IT is required to obtain optimized energy systems and to enable load control in a deregulated electricity market where price comparisons determine whether equipment shall be disconnected or not.

For example, for distributors, who wish to operate at their margins for income maximization, the exact amount of load needs to be controlled against these limits. For optimized conditions, the loads of the lowest management prices are to be utilized first in a the management procedure. And most important, the exact amount of power needs to be controlled. Thus, price signals need to be weighed against each other and only the loads deactivated that lead to an entire system optimization. 64 Theoretically, one-way communication can satisfy the requirement of price comparison and would, therefore, also be enough to enable the energy system optimization. Real time price signals to customers would eventually be responded with direct load disconnection. The customer response, i. e. The load reduction would be registrated by the distributor in critical meter points of the distribution grid. Another option is that electricity contracts would regulate the availability of load control against electricity price. Above a certain price level, the distributor is free to disconnect the equipment, and below this level, no regulations are allowed.

Thus, energy system optimization requires one-way communication of price signals and simultaneous registration of electricity consumption. One solution is to distribute directly these meter values back to the supplier/distributor (two-way communication), but they canalso be registered and, for example, manually collected by the distributor.

On the other hand, part of the customers themselves could supply the distributor with relevant information for eventual load control. For example, industries with process- related equipment, where the price for disconnection varies in duration, could supply the information to the distributor in order to let him optimize the assets of the entire system. The advantage with such a system is that the time-consuming iteration procedure required to obtain exact load control, for example, at the margin of the distribution capacity, could be replaced by a relatively fast optimization process performed by the distributor. Price signal only, transmitted to the customers, is likely to bring too high or too low amount of power disconnected.

However, most of the solutions found in literature recommend two-way communication for system optimization at a deregulated electricity market. One example is the by Ygge [86] described application of market oriented programming to enable load control. Briefly, a computational agent, a HOMEBOT, is here representing each load. Its task is to competitively maximize its utility through the trade with other HOMEBOTS - two- way communication. For other examples see section 3.1.2.

Thus, at a deregulated market, it is likely that two-way communication would be required to form conditions acceptable for both the distributor and the customer. If nothing else, only as a symbol of the radical shift of power.

5.5.8 Market actors incentives and financing Studying the cost savings obtained from load control of residential customers in Ronneby and comparing those with the cost for PLT, it is evident that these savings scarcely can finance the equipment required to enable the control (see the enclosed Paper I). Approximately 900 000 SEK divided among 3 300 customers means around 270 SEK/year to be invested in communication equipment. Experience from the ID AM project points at investment costs that farexceed this amount (see section3.5.4).

65 Similar experiences are reported by Andersson [1], who regarding smart house services 41, claims that the cost for the enabling IT infrastructure and equipment are high, compared to theeconomic benefits expected in the short term.

But, because of their numbers, the small customer market is comparable in size to that of the larger customers, and, therefore, it is also still important regarding load control to obtain system optimization. However, for the moment, with the current interaction technology, the main difference is the cost of interaction with the many instead of the few.

Focusing on the market player incentives for load management, in general, these are also rather few. Since few distribution limitations are to be found in the Swedish distribution grid, most distributors have little interest in such measures. This is a consequence of the political effort to expand electric heating in the 1970s. With no need to operate on the limits of the distribution capacity, due to tariff construction, load control mainly reduces the income of the grid owner. By reducing the power peaks of his customers, he also reduces his income from power-related charges. These charges are generally higher than those the distributor himself paysto the regional net operator.

However, looking at local energy system optimization, to minimize the energy system cost of a distribution concession area, load management should be introduced in cooperation with the customers to attain lower power-related charges to the regional net operator. The more manageable the equipment that are available for the distributor, the lower subscription agreement he can make with the regional distribution company, which means thathe will pay a lower fee (see theenclosed Paper I).

OPTIMAL SOLUTIONS DEMAND CURVE

SUPPLY CURVE

QUANTITY

Figure 5.12 System cost reduction due to increased load control illustrated in a supply demand graph

41 Smart house services generally include load control, see section 3.4.2. 66 Neither do electricity sales companies have a great interest in the reduction of power peaks for short-term yieldmaximization. As no power-related charges exist at the power exchange and their exposure to the prices of the Regulating Exchange is rather small, theirmain reason for load control introduction is to retain or to win customers.

Even when exposed to the prices of the Regulate Exchange market, load control is of little concern. The prices here seldom exceed 400 SEK/MWh. Looking at Ronneby, at this price level, the sales company, Blekinge Energi, would at the most be able to disconnect around 11 MW within Ronneby. Each hour thecontrol was utilized, it would, therefore, lead to a savings of around 4 400 SEK. On the other hand, it would also lead to a loss of income due to the reduced electricity sales. Since the sales companies normally try to avoid the regulate exchange, such situations might occur around 20 times within a year. Annually, the control would, therefore, correspond to a savings of around 90 kSEK, minus the income from electricity sales. Provided that the power is sold for around 200 SEK/MWh; the total saving would only be 44 kSEK annually. Divided on the residential customers, it leaves an investment capacity for communication equipment of less than 15 SEK/customer.

Sometimes though, the regulate price becomes very high, around 1 SEK/kWh. At these very seldom occasions, load control of course is a lot more beneficial for the sales companies. 8 800 SEK would be saved by the sales company at each hour the control was utilized. But, these extreme price levels very seldom occur. If Blekinge Energi was exposed to these levels five times in one year, their annual savings due to the control would still only become 44 kSEK. Therefore, exposition to the Regulation Exchange cannot finance large scale PLT based load control of residential equipment. Further, the level of the power charges of the monopoly distribution contracts are much higher (around 105 SEK/kW in 1997), which leads to customers concentrating on reducing this charge.

Another possibility would be that the sales companies, in times of high electricity purchase prices, would like to disconnect customers who have electricity prices below the purchase level. Provided that this was the case in Ronneby and thatall the concerned customers had a price of 200 SEK/MWh when the spot purchase price was 250 SEK/MWh, each hour Blekinge Energi disconnected these customers ’ equipment, the company would earn 550 SEK. Assuming that the purchase price reached 250 SEK/MWh, half the time of December, January and February (1008 hours), the savings from load control would be around 550 kSEK. Still, divided among the customers, it leaves approximately 160 SEK/customer to investment. To give a comparison, in 1997, the diurnal average prices exceeded 200 SEK/MWh for only 28 days; in 7 of these, the price exceeded 250 SEK/MWh. The number of hours interesting for load control was, therefore, around 670 hours totally, and only 170 hours had a price level higher than 250 SEK/MWh.

Small electricity producers could theoretically have a corresponding interest in load control, since they might like to disconnect unprofitable customers in favour of selling capacity at higher price levels. However, the economy regarding the investments is similar to thatof the sales companies. 67

1 mSI Important is also that this possibility is only open to smaller producers. Large-scale actions performed by a dominating producer operating at the exchange are likely to affect the price level, which, as discussed in section 5.5.6, follows the demand with a slight delay.

Together, this means that given the boundary conditions of the current electricity market, small incentives are given to operating actors to seek DSM solutions optimal for thesystem in its entirety.

Instead, the benefit from load control can mainly be found by the industrial customers, see the enclosed Paper II. But, given the current electricity contracts, these customers will mainly utilize load priority system to control that theirtotal load does not reach the peak power level agreed upon in the distribution contract.

Further, increased introduction of new production philosophies, such as, ’’Just in Time” (JIT) and Total Quality Management (TQM), are likely though to reduce the availability of manageable process-related equipment within industries. Another consequence is that units still available probably will become more expensive for the distributor to disconnect than earlier, provided that the cost corresponds to their importance for the production process. Thus, initiatives for system optimization is left to the distributor only.

Regarding the investments in IT infrastructure and hardware, the electricity prices and generation costs of today cannot, as shown, defend investment in IT applications for large-scale load control of residential equipment since the costs for remote metering and two-way communication are still too high. However, the electricity sales driven introduction of communication-based VAS, discussed in section 3.4, will most probably, as a side effect, lead to more effectively utilized energy systems. Load control applications are here likely to piggy-back on the costs of other more attractive offers for the customers, like Internet access or security services. The electricity sales companies ’ concentration on IT-related VAS and their view of two-way communication as a strategic link to the customers are therefore, likely to also bring the customers increased power and energy-regulating control possibilities.

Therefore, regarding financing, two-way communication is likely to become the applied technology for load control, even if one-way communication theoretically could enable the energy system optimization.

6. COMMENTS ON ENCLOSED PAPERS

Paper I : Even though this paper was published after the Swedish deregulation, its conclusions are partly drawn from the perspective of an entire local energy system optimization, including both electricity sales and distribution. In such an optimization, the different market actor incentives for certain measures are not considered. Instead, in a system perspective, focus is laid upon how to minimize the total cost for supplying, in 68 this case Ronneby, with heat and power. For instance, in the analyses, no consideration is taken of the fact that the utility, READ, after the deregulation, only handles electricity distribution, and not the purchase or sale of electricity. The latter is, instead, taken care of by the regional electricity sales company, Blekinge Energi. Therefore, regarding the conclusions of benefits from DSM, these are also true, based on the traditional view of electricity markets and local market cooperation.

But, focusing on the regional sales company only, load control is of little interest to them, as said in the paper, mainly due to the lack of power-related charges on the spot market. Even exposed to the prices of the Regulate Exchange market, which seldom exceed 400 SEK/MWh, load control is of little concern (see section 5.5.8).

However, also sales companies may make savings from load control when the timely variations of their purchase costs are larger than the timely variations of the electricity sales prices. In this paper, this is illustrated by the economic benefit from the load control of residential customers. Unfortunately, the earnings for the single customer is rather small, namely, 73 SEK annually.

Viewing the distributor, REAB, their savings from energy transpositions are marginal, as the distribution prices in 1997 varied between 2.0 and 1.1 SEK/MWh. Regarding load control against the power fee, REAB would earn from the disconnection of residential equipment, since these customers normally have no possibility to reduce their power charges due to the tariff construction. Thus, only the regional power charge paid by REAB, would be reduced. Further, load control is, of course, always favourable when the regional charge is higher than the power-related charge paid by the customer. If that is not the case, the distributor would mainly lose money due to thecontrol.

For the energy system though, seen as one unit, DSM measures are always most interesting, since these actions can considerably reduce the total cost of supplying the entire system.

Regarding, investment in a new production capacity the analyses show that combined heat and power would reduce the energy system costs by roughly 30 MSEK in five years. The figure 6.1 shows how such a plant would be operated.

69 CHP IN P#NNEBY LPG hellers "V-- Electric hellers

Figure 6.1 Duration graph of heat demand and potential optimized supplyin Ronneby

Paper II: In paper II, the price increase of local power-related distribution charges is made evident. Thus, load control has become more profitable for industrial customers, but less beneficial for the distributor. The power charges are sometimes increased by almost 90%. In contrast, between 1997 and 1998, the regional distribution charge to Ronneby has only been increased from 12+105 SEK/kW to 13 + 105 SEK/kW. The first figure corresponds to the annual power fee, and the latter to an additional charge, in order to account for the winter week days.

Since many local distribution companies have increased the customers ’ power fee after the deregulation, this charge has become considerably higher than the regional fee paid by the grid operator. Thus, load control causes an even larger loss of income for the distributor after the deregulation than it did before. Therefore, the distributor incentives to introduce load control, are mainly due to limitations in the distribution grid. However, provided that the electricity sales company, or the customers themselves, plan to realize load priority systems or similar applications, the distribution company could partly reduce its economic losses, through directed distribution tariffs to these customers. This is also exemplified in this paper, even though directed distribution tariffs are so far illegal.

Paper III: In this paper, the consequences on the distributor are shown if real time pricing, or spot prices, are introduced only to part of the customers and if no power- related charges influence the variations of demand. Provided that all customers would

70 get the same price offers, the variation of price is likely to follow the variation of demand.

It is further shown that the power charges determine when and how eventual load control should be utilized. Even though the construction of power-related charges does not mirror the capacity of the distribution grid (see 5.5.5), when load control is utilized, these charges do lead to more flattened load profiles, which is an advantage in energy system operation (see section 2.3).

An important consequence of real time pricing is also that load control might be more utilized during the year, due to permanent price variations. Though optimization models, can only scantily describe these measures, since future prices, in reality, are neither known by the customers nor by the suppliers. Therefore, it is more likely that load is disconnected when prices exceed a certain level, already agreed upon.

Regarding the supply-demand graph, it illustrates how Nord Pool determines the spot price of the coming day, due to the power balances delivered from the spot market participants. In macro-economic theory, the demand function needs to be a continuous curve.

7. CONCLUSIONS Briefly, this thesis concludes that information technology is essential for obtaining optimized energy systems. It further points out that the utilization of IT within electricity sales companies has increased, both in the strategic area of Value Added Services and for administrative issues. Electricity sales companies, thereby are likely to become dependent on information technology.

Results from energy system optimizations have shown that the ability to control load is of vital importance to obtain optimized energy systems (see the enclosed Paper I). The analyses further point out a considerable reduction of the energy system cost, which occurs when, apart from industrial energy-related equipment, also residential equipment is included in the load control. Industrial simulations have similarly proven the benefits from additional control of process-related equipment (see the enclosed Paper II).

To make the most controllable load available as possible, the experiences from the industrial simulations (see section 5.5.4.), together with studies of IT-related pilot projects (see section 3.5) point out the need for information technology to:

1) enable the load control in practice. 2) ensure that disconnection of equipment is performed according to current prices and agreements. 3) ensure that the control does not cause inconvenience or economic failure for the parties involved.

For similar reasons, IT is required to enable load control of residential equipment.

71

% % 'J' Diurnal energy system analyses have further shown the increased utilization of load control when customers are exposed to the variations of the spot market price (see the enclosed Paper III). This illustrates the increased need for load management, on a deregulated market where real time pricing is applied. In accordance with the assumptions regarding optimized energy systems (see section 5.5.1), to obtain such systems, sudden changes of SRMC of electricity production need to be compensated by load control (see section 5.5.6).

Thus, as the time division of electricitypricing is refined and price variations sometimes are considerable, the need for increased flexibility in load rescheduling becomes urgent, both on a system level and viewed from the customers ’ perspective (see the enclosed Papers II and III).

However, given the conditions of the deregulated electricity market, it is further shown that these do not necessarily lead to more optimized energy systems. As overviewed in sections 5.5.3 and 5.5.8, the market actor incentives for practical realization of large- scale load control are rather few in comparison to the overall potential of energy system optimization to reduce the energy systemcost. Distributors mainly lose money on power reducing measures, except at distribution limitations. Producers have little interest in residential load rescheduling due to high corresponding costs compared to the potential savings. Neither electricity sales companies make short-term earnings from large-scale load control, still, they are likely to introduce load control as part of their VAS strategy (see section 2.2).

In the attached Paper I, and in comments on the enclosed Papers I, it is further shown that the control of residential hot water heaters and dual fuel boilers only brings low savings for the customers at current electricity price levels. In contrast, large industries may gain considerable amounts through load rescheduling (see Paper II). The financing of the latter solutions comes natural, but, unfortunately, IT-based solutions to enable large-scale load management of residential customers is more problematic (see sections 3.5.4 and 5.5.8).

Regarding IT utilization within the power industry in general, compilation of interviews and literature surveys point at IT as one key parameter that will influence the future power business. In the areas of power system balancing, metering, distribution maintenance and electricity sales, IT is the common factor, applied for totally different objectives (see sections 3.1, 3.2, 3.3 and 3.4). Generally, the strive towards increased utilization of existing infrastructure assets lead to increased IT utilization within many business operations in the power industry.

It is further shown that electricity sales companies have increased their utilization of the Internet for informative reasons, for offers and for administrative services (see section 3.4.2). It is also pointed out that large parts of their strategic actions already taken, or about to be realized, are IT-based.

72 Increased IT utilization by electricity sales companies is observed to enable services within:

• marketing and service offering on the Internet • customer administration, customer service and • IT-based VAS • pure communication services, like telephone and Internet access

Viewing the strategic development of VAS, IT is likely to become further utilized for coordination of VAS, based on value networks. Another also probable development is introduction of smart house services (see sections 3.4.1 and 3.4.2).

Finally, regarding financing of large-scale load management, as mentioned above, currently, this cannot be achieved by the benefits from load control itself, but rather through the piggy-back on the costs of other information flows, like Internet access and security services (see section 5.5.8). Thus, the boundary conditions of the deregulated market may not favour optimization of energy systems through load management, yet, the development of IT-related VAS is likely to, as a side effect, in the future, enable increased utilization of load control.

8. ABBREVIATIONS

AMS Asset and Maintenance Management System BMS Business Management System BP Back Pressure plant CHP Combined Heat and Power plant CIS Customer Information System CMMS Computerized Maintenance Management Systems DA Distribution Automation DMS Distribution Management System DNO Distribution Network Operator DPL Digital Power Line DSM Demand Side Management EDI Electronic Data Interchange EMS Energy Management System ESC Electricity Sales Company FACTS Flexible AC Transmission System CIS Geographical Information System GT Gas Turbine HWP Hot Water Plant IA Intelligent Agents IT Information Technology LAN Local Area Network LC Load Control LPG Liquid Petroleum Gas LRMC Long Range Marginal Cost MO Meter Operator NOK Norwegian Crowns 73 PDC Producer Distributor Customer PLT Power Line Telecommunication RCM Reliability Centred Maintenance RTP Real Time Pricing SCADA Supervisor)' Control and Data Acquisition SEK Swedish Crowns SRMC Short Range Marginal Cost VAS Value Added Services WAN Wide Area Network

9. SELECTED REFERENCES

1. Andersson Jonas, On IT System Integration, Prospects and Consequences of Energy Market Deregulation, Licentiate Thesis 1997, Industrial Control Systems, School of Electrical Engineering and Information Technology, KTH, Sweden

2. Andersson M, Cooperation in local electricity markets: Modelling of Technical Measures, Linkoping Studies in Science and Technology Dissertations No. 470, 1997, Linkoping Institute of Technology, Sweden

3. Andersson M., Cost-effective Incentives for Local Electric Utilities and Industries in Cooperation, Modelling of Technical Measures, LiU-TEK-LIC 1993:12, Department of Mechanical Engineering, Linkoping Institute of Technology, Sweden

4. Andersson Per, Nya produkter och tjanster - forberedelser infor avregleringen av elmarkanden for hushallskunder, Statt Utlandsrapport 1997, Sveriges Tekniska Attacheer, Box 5282, 102 46 Stockholm, Sweden

5. Akkermans H, Healey D, Ottosson H, The transmission of Data over the Electricity Power Lines, 1998, Published in The ISES Project, ISBN 91-9753567-0-0s, pp 175-216

6. Akkermans H, Ottosson H, Ygge F; The ISES Project, Information/Society/Energy/System 1996-1998, ISBN 91-9753567-0-0s

7. Augustsson Kamilla, Utveckling av tjanster p& elmarknaden i USA, Statt Utlandsrapport 1997, Sveriges Tekniska Attacheer, Box 5282,102 46 Stockholm, Sweden

8. Backlund Lennart, Optimization of Dynamic Energy Systems with Time Dependent Components and Boundary Conditions, Linkoping Studies in Science and Technology, Dissertations No 181, Linkoping Institute of technology, Sweden, 1988

9. Baderschneider C., Larsson R., Sweet P., High-Precision Energy Marketing: Customer Style Accomodation thgrough Sales Partnering and Interactive Market Learning, In the proceedings of the Distribution Automation Demand Side Management Conference, Amsterdam 1997, The Netherlands

10. Bergstrom U., Nilsson P., Forstudie av biobransle eldad hetvatten produktion respektive kraftvarmeproduktion i Ronneby, Sydkraft Konsult AB, 1997,205 09 Malmd

11. Bergstrom, U., Impact of deregulation - DSM in industries, Proceedings of the international conference on Distribution Automation & Demand Side Management 74 (DA/DSM), CD ROM, London, UK, Penwell Conference and Exhibitions, P.O. Box 9402, 3506 GK Utrecht,The Netherlands, October, 1998.

12. Bergstrom, U., Simulation of a local energy system with focus on cost efficient DSM measures on a deregulated electricity market, In Proceedings of the international conference on Distribution Automation & Demand Side Management (DA/DSM) Europe Conf. Vol. Ill, Vienna, Austria, Penwell Conference and Exhibitions, P.O. Box 9402, 3506 GK Utrecht, The Netherlands, October, pp 631-644,1996.

13. Bergstrom, U., Traditional load management versus real time load control, Presented for the European RISI conference, November, Graz, Austria 1998.

14. Bj6rk Curt 0, Industrial Load Management - Theory, Practice and Simulations, Energy Research7, Linkoping Institute of Technology, Linkoping, Sweden, 1989

15. Brennan P, Welke A, Thornton R, Integration of SCADA and DA/DMS across a Large Distribution System, South West Power & Introl, Proceedings of the International Conference on Energy Management and Power Delivery, EMPD, Nov 21-23 1995 v 1 1995 Singapore, IEEE Catalouge No 95TH8130, 1995 pp 31-36

16. Brorsson L., Sondh B. T., Sydkraft AB, personal communication 1998 regarding IT utilization in distribution supervision.

17. Cheong, Kam-Hoong, Distribution Automation - Cost-Effective Introduction Strategies, Licentiate Thesis 1997, Industrial Control Systems, School of Electrical Engineering and Information Technology, KTH, Sweden

18. Dahlfors F, Pilling J, Integrated information Systems in a privatized and dergulated electricity market, Proceedings of the International Conference on Energy Management and Power Delivery, EMPD, Nov 21-23 1995 v 1 1995 Singapore, IEEE Catalouge No 95TH8130, 1995 pp 249-254

19. Effektstyming p& hog niva, Energimagasinet 2:95

20. Elmarknadema i Europa 1992, NUTEK, 11786 Stockholm, Sverige

21. Energilaget 1997, NUTEK, 11786 Stockholm, Sverige

22. Energitidningen 3.98, Vattenfall

23. Elnytt 2, Nyhetsbrev fran Sveriges Tekniska Attacheer, Los-Angeles och London, 1998, Sveriges Tekniska Attacheer, Box 5282,102 46 Stockholm, Sweden

24. Fahlgren B, Handel med el - kunders beteende pa en konkurrensutsatt elmarknad, Statt Utlandsrapport, 1995, Sveriges Tekniska Attacheer, Box 5282,102 46 Stockholm, Sweden

25. Flatabo N., Information Technology related to challanges in a liberalised power market, In the Proc. Of the second IEEE International Caracas conference on Devices, Curcuits and Systems, Isla de Margarita, Venezuela, 1998

26. Forsberg T., Sydkraft AB, personal communication 1998 concerning IT utilization in power production optimization and physical exchange.

75 27. Goldman C, Impact of Information and Communications Technology on Residental Customer Services, Lawrence Berkley National Laboratory, University of California, 1996

28. Goshal K, Distribution Automation: SCADA Integration is Key, IEEE Computer Applications in Power, New York, January 1997, ISSN 0895-0156, pp 31-35

29. Gronli H., Lillefloth R., Prisfolsamhet overfor tidsvariable tariffer, arbetsdokument, 1996, EFI SINTEF Gruppen, Energiforsyningens Forskningsinstitutt AS, 7034 Trondheim, Norge

30. Half-Yearly Report 1, 2 1996, UCPTE, 2, rue Pierre d ’Aspelt - BP37 - L-2010 Luxembourg

31. Henning Dag, Energy Systems Optimization Applied to Local Swedish Utilities, Linkoping Institute of Technology, Division of Energy Systems, Department of Mechanical Engineering, Linkoping, Sweden, 1994.

32. Henning D, Optimization of Local and National Energy Systems - Development and the Use of the MODEST Model, Dissertations No.559, Division of Energy Systems, Department of Mechanical Engineering, Linkoping University Sweden 1999.

33. Hirschleifer, J., Price Theory and Applications, fourth edition, Pentice Hall International Editions, ISBN 0-13-698334-0, USA,1998

34. Hoffman S., Renner R., Taking Advantage of Real-Time Pricing, The EPRI Journal, March/April 1997, Vol 22, Number 2: pp 16-23

35. IDA Rapport 1, Sydkraft internal report 1995, Sydkraft AB, 205 09 Malmo, Sweden

36. Jansson L, Sydkraft Electricity Sales, personal communication 1998 regarding future IT- related services

37. Johansson K., Norsk Hydro Syd AB, personal communication 1998 regarding Norsk Hydro electricity sales strategies.

38. Johansson K, Sydkraft Distribution, personal communication 1998 regarding IT in electricity distribution.

39. Jonsson M, Sydkraft Electricity Sales, personal communication 1998 concerning Internet marketed VAS.

40. Keijer U, Ett individperspektiv for det nya IT-baserade servicesamhallet Roller for Energiforetag i nya affarsstrukturer, Institutionen for Form och Teknik, KTH

41. Keijer U, Nilsson B, Technology, Organisation and Market - The Application of a Model for Concurrent Business Development for the Emerging IT-based Home Services Market, The 6:th conference of the Joseph A. Schumpeter Society on Competiotion, Entry and Economic Growth - The Firm, the Innovator, The Entrepreneur and Market Competiotion Stockholm 1996

42. Koponen P., Remote reading and data management system for kWh-meters with power quality monitoring, International Conference on Metering and Tariffs for Energy Supply, Bright, 1996, Conference Publications No. 426, IEE, pp 11-15 76 43. Kraftaret 1997, Svenska Kraftverksforeningen, 101 53 Stockholm, Sverige

44. Kreutzberg, M., EUDIS Forecasting Spot Prices for the European Power Market, Institute of Energy Economics University of Colonge, November 1998

45. Lamarre L, At home with Telecommunications, EPRI Journal, Jan/Feb 97.

46. Larsson L, SYCON Energikonsult, personal communication 1997 and 1998 regarding electricity pricing.

47. Larsson R., Passby M., High-Precision Marketing through Customer Style Accommodation: Utility Industry Applications, In the proceedings of the Distribution Automation Demand Side Management Conference, Amsterdam 1997, The Netherlands

48. Larsson R., Sweet P., Interactive Marketing Information Systems: Towards High-Precision Market Communication through Electronic Media, In the proceedings of the Distribution Automation Demand Side Management Conference, Vienna 1996, Austria, pp 535-550

49. Lethonen M, Kupari S, A method for cost benefit analysis of distribution automation, Proceedings of the International Conference on Energy Management and Power Delivery, EMPD, Nov 21-23 1995 v 1 1995 Singapore, IEEE Catalouge No 95TH8130, 1995 pp 49- 54

50. Lethonen M, Seppela A, Kekkonen V, Koponen P, Koreneff G, Lemstrom B, Distribution load estimation - A function for the electric utility in the deregulated electricity market, Proceedings of the 1995 International Conference on Energy Management and Power Delivery, EMPD'95. Part 1 (of 2) Nov 21-23 1995 v 1 1995 Singapore, pp 240-243

51. Lundgren P, Sydkraft Electricity Sales, personal communication 1998 concerning IT- related value added services.

52. Lof P-A, Styming och Overvakning av Elkraftsystem i Nordamerika, Statt Utlandsrapport, maj 1998, Sveriges Tekniska Attacheer, Box 5282,102 46 Stockholm, Sweden

53. Makansi, J., 1995 Engineering Information, Inc., Hoboken, New Jersey. USA, Record 30 : Information Technology for Powerplant Management, Power v 139 n 6 Jun 1995 McGraw- Hill, Inc. New York NY USA 8pp ISSN: 0032-5929 CODEN: POWEAD

54. Marzio L, A new utility-user interface for a qualified Energy Consumption, Author Affiliation: Enel, Source: Pattern Recognition v 28 n 10 Oct 1995, Pergamon Press Inc TarrytownNY USA p 1507-1515, ISSN: 0031-3203

55. Mccartney A, EMS Project Overview, Proceedings of the IEE Colloquim on the New NIE Energy Management System, May 1998, ISSN 0963-3308

56. Mansson Cecilia, El av hog kvalitet i USA - en tjanst med mervarde eller en vinstgivande produkt?, Statt Utlandsrapport 1994, Sveriges Tekniska Attacheer, Box 5282, 102 46 Stockholm, Sweden

57. Nara K, Hasegawa J, An advanced flexible and reliable distribution system, Proceedings of the International Conference on Energy Management and Power Delivery, EMPD, Nov 21- 23 1995 v 1 1995 Singapore, IEEE Catalouge No 95TH8130, 1995 pp 55-60 77 58. Nilsson K.,Cost-effective Industrial Energy Systems, Multiperiod Optimization of Operating Strategies and Structural Choices, Linkoping Studies in Science and Technology, Dissertations No. 315, 1993, Department of Mechanical Engineering, Linkoping University, Sweden

59. Nilsson P.A., Scott D., Competitiveness of future coal fired units in different countries, 1999, IEA Coal Research, Publications Department, Gemini House, 10-18 Putney Hill, London SW15 6AA, UK

60. Nilsson R, Nord Pool, personal communication 1997 and 1998 concerning the power exchange.

61. NORDEL Arsberattelse 1997, Imatran Voima Oy, Malmg 16, Helsingfors

62. Norman R., Ramirez R., Designing Interactive Strategy: from value chain to value constellation, 1998, ISBN 0-471-98607-0, pp 33,45-65

63. Nyman J, Nya affarsmojligheter med Intelligenta Hus, Statt Utlandsrapport 1998, Sveriges Tekniska Attacheer, Box 5282,102 46 Stockholm, Sweden

64. Naringsutskottets betankande 1997/98:NU3 ,Ny ellag

65. Olausson H, president Ronneby Energy AB, personal communication 1996-1998, regarding the development of the Ronneby energy market.

66. Optimal Control: A Typical Day, The EPRI Journal, March/April 1997, Vol 22, Number 2: PP 21

67. O'Sullivan John B, Application Experiences withDistributed Resources in the Midwest, Proceedings of the American Power Conference 1998, Part 1 (of 2) v 1, Chicago, IL, USA 2 Illinois, pp 47-57 ISSN: 0097-2126

68. Persson Maria, IT-anvandningen i elbolag - idag och imorgon, STATT Utlandsrapport 98, Sveriges Tekniska Attacheer, Box 5282,102 46 Stockholm, Sweden

69. Persson B, Sydkraft AB, personal communication 1998 concerning IT applications for deduction.

70. Persson, Thomas, Distribution Automation - A Utility Perspective, a paper for the I EE seminar ’’Methods of Substation Automation ” 30 May 1996, Author Affiliation: Malmo Energi AB, 205 09 Malmo, Sweden

71. Power and the Future, The EPRI Journal, January/February 1998, Volume 23, Number 1, pp 16-25

72. Product specifications of Sydkraft Value added services, 1998, Sydkraft Elforsaljning AB, 205 09 Malmo, Sweden

73. Prognos 1995, (in Swedish), The Swedish District Heating Association, Stockholm, Sweden 1995

78 74. P&lsson T, Elektronoisk handel over internet II - de nya affarsmodellema, Statt Utlandsrapport 1998, Sveriges Tekniska Attacheer, Box 5282, 102 46 Stockholm, Sweden

75. Schulte R, IT Architecture vs. IT City Planning, GartnerGroup Symposium ITexpo98, The future of IT. Cannes, France, Nov 1998

76. Smart telefonjack Erikssons framtid, Dagens Industri, February 19,1999

77. Stanway J, The Holyhead Experience, Power Engineering Journal, October 1994

78. Stoa P, Soft Communicating Meters, International Conference on Metering and Tariffs for Energy Supply, Bright, 1996, Conference Publications No. 426, IEE, pp 149-153

79. Strukturforandringar p& den svenska elmarkanden, Sveriges El Leverantorer 1998, 101 53 Stockholm, Sweden

80. Teknikstyming av elanvandning rapport 1, (Technical control of electricity usage report 1, in Swedish), SEU, Sydkraft 205 09 Malmd, Sweden, 1989

81. Utvecklingen p& elmarknaden Oktober 1996, NUTEK, 11786 Stockholm, Sverige

82. Verho P, Jarventausta P, Karenlampi M, Partanen J, Intelligent Configuration Management of Distribution Network, Proceedings of the International Conference on Energy Management and Power Delivery, EM PD, Nov 21-23 1995 v 1 1995 Singapore, IEEE Catalouge No 95TH8130, 1995 pp 43-48

83. Weiner M, Nohria N, Hickman, A, Smith H, Value Networks - The Future of the U.S Electric Utility Industry, Sloan Management Review, Summer 1997

84. Wildberger Martin A, Modeling with independent intelligent agents for distributed control of the electric power grid, Proceedings of the American power conference/ Chicago, Illinois, 1997, pp 361-364

85. Williams BR, Distribution Capacitor Automation Provides Integrated Control of Customer Voltage Levels and Distribution Reactive Power Flow, Power Industry Computer Application Conference 19, Salt Lake City 1995

86. Ygge F., Market-Oriented Programming and its Application to Power Load Management, Dissertation, Department of Computer Science, Lund University, 1998

87. Ostholm, Svante, Behovet av strategisk fomyelse, Innovation.nu, 1998, Klaraberg Managementkonsulter AB, Stockholm

Web-addresses at Internet http://www.alingsas.se/kommun/energi http://www.bergen-energi.com http://www.boden.se/beab http://www.boras-energi.com http://www.booenergi.se http://www.borgholm.se/borgholmenergi http://www.gavleenergi.se http://www.eastem,co.uk 79 http://www.energi.goteborg.se http://www.energi-miIjo.kalmar.se http://www.energy-save.com http://www.eme.co.uk httpV/www.eskilstuna.se/energi&miljo http://www.faluenergi.se http://www.gka.se http://www.graninge.se http://www.helsingborgenergi.se http://www.hemab.se http://www.hemel.se http://www.hjo.se/ftg/hjoenergi http://www.hke.fi/hg(Helsingfors Energi) http://www.jobbel.se http://www.kraftoenergi.com (Billeberga Kraft o Energi) http://www.kvanumenergi.se http://www.london-electricity.co.uk http ://www. lundsenergi . se http://www.nie.co.uk http://www.nordpool.no http://www.northem-electric.co.uk http://www.nynasenergi.se http://www.oslo_energi.no http://www.ostfoldenergi.com http://www.ronneby.se/reab http://www.sandnes-energi.no http://www.scottishpower.plc.uk http://www.seab.se http://www.seeboard.co.uk http://www5.sydkraft.se/ne (Norrkopings Energi) http://www.svk.se (Svenska Kraftnat) http://www.sweb.co.uk http://www.tekniskaverken.linkoping.se http://www.telgeenergi.se http://www.vattenfall.se http://www.veab.se http://www.yeg.co.uk http://www.ystad.se/text/service/teknisk_service/ye_allment

80 SIMULATION OF A LOCAL ENERGY

SYSTEM WITH FOCUS ON

COST EFFICIENT DSM MEASURES

ON A DEREGULATED

ELECTRICITY MARKET

Ulrika Bergstrom

Sydkraft Konsult AB

and

the University of Linkoping

81 Summary Could a utility gain anything from spending money on demand side management? The answer is yes, and espescially with the diurnal price differentation and the power related charges of todays Swedish electricity delivery contracts. The analysis of the Ronneby energy system shows that the combination between load management and efficiency improvments colud reduce the costs of supplying Ronneby with electricity and heat with 12 Million Swedish crownes (MSEK), real value, in 5 years, 0% electricity price raise and all investment costs included. This means a reduction of 2,9% of the total cost for supplying the system. Load management it self could reduce the cost with 7 MSEK in the same time period, investment costs included. About 5,6 MW electricity would then be controlled in the residential sector during the winter period and roughly 4 MW in the industry at the same time.

Load management will though probably be of less importance at the future Nordic electricity market. Todays absence of any power related charges at the diurnal power exchange market does not favour reduction of power peaks which makes load management less profitable. The deregulation will also in the long term generate a more stable price level of electricity at the diurnal power exchange market, with smaller price variations between high and low load periods. Neither this favours load management. The existing intemiptable equipment of the energy system will in the future more likely be used for energy transpositions instead of for the cutting of power demand peaks as the regional and local electricity sales companies start to act at the diurnal market. Analysis of the Ronneby energy system shows that introduction of load management when the electricity is purchased on the diurnal power exchange market would reduce the system cost with less than 1% in 5 years.

The electricity market in Sweden, Norway and Finland is today roughly 325 TWh. It is forecasted to increase slightly more than 1 %/year which corresponds to about 4 TWh/year in the nearest future. Due to long pay-off times the deregulation does not favour big investments in new electricity production capacity. Therefore, small scale energy technology and combined heat and power producing units (CHP) are interesting investment alternatives in the future. Sweden has a unique demand for both electricity and heat at the same time, which promotes CHP.

Introduction of CHP (4 MWe/12 MWh) in Ronneby would reduce the system cost about 6,5%, (28 MSEK), at current conditions, and 7,2% if electricity was purchased at the power exchange market, 0% electricity price raise and investments costs included. Together CHP and DSM could reduce the system cost with up to 12%, 50 MSEK in 5 years, at current electricity contracts.

Both CHP and DSM measures are powerful tools at the deregulated market. With CHP and a large capacity accumulator the distributor can choose at what time he wants to buy electricity or to produce it himself. Load management gives him the opportunity to control a part of the electricity demand, and with efficiency improvements, he shows that he cares about the energy costs of his customers. In a competitive market it is likely that if he does not care, somebody else will and many of the customers will of course prefer the energy supplier who does.

Introduction As part of the ISES project (Information, Society, Energy, System), a project for introduction of IT-solutions in the Blekinge region, the analysis here is focused on the profitability of DSM in the energy system of Ronneby. Since the deregulation of the Swedish electricity market took place and competition occurred at the electricity market, many electricity sales companies try to bring so called "add values" to their product to make it even more attractive to the customers. One of those "extras"

82 could be different IT services, as for example smart houses and load management. Through information technology it would be possible for the electricity suppliers to communicate with different equipment at thecustomers and that way both reduce the customers ’ energy costs and increase their comfort. In the vision, the communication needed for these services, would mainly take place on the electricity grid. Today this is already possible in a smaller scale.

But as in all market driven societies, made investments have to pay and the analysis presented here has been made to elucidate the economical incentives, for or against demand side management in the Ronneby energy system.

Methodology For analysing the energy system the linear programming model MODEST, originally developed at the University of Linktiping, has been applied. With the model it is possible to almost perfectly describe any energy system in economical and technical terms (see Model Description). At first though it is important to define the system limits for the analysis and make clear which actor that shall be focused on. For example, if, as in this case, the benefits of the electricity distributor is focused, energy taxes, that the distributor has to pay to his supplier and that he also allocates to his own customers, shall not be included in the analysis. Concerning electricity prices, written contracts between the supplier and the distributor shall be analysed, and changes of income from the distributors electricity sales must as well be considered. If instead the producer is focused the electricity prices analysed, must be his own marginal electricity production costs and the prices of the power exchange market. The conclusion is that the measures recommended in an analysis varies considerably depending on the system limits and which actor that is focused.

The definition of system limits is followed by a considerable collection of input data and a period of careful re-checking. Investments costs as well as price trends for different energy carriers are to be estimated and running schemes for the analysis have to be made. A delicate problem that has occurred in the analysis after the deregulation of the electricity market is that it is no longer possible to predict a distributor ’s future demand profile for electricity. This fact makes it even more important to make careful forecasts for electricity prices. Because if there is a strong price raise for electricity, CHP production units will be profitable due to the high electricity pricing, and will not be limited due to the demand profile for electricity or to made agreements with electricity suppliers. Heat storages will as well probably charge as much as possible during daytime when the electricity prices are high and discharge during the night when the electricity prices are lower etc.

Forecasts At the future electricity market most regional and local electricity sales companies will probably ask fore relatively short electricity delivery contracts that both allow them to buy from one or two main suppliers and to purchase on the power exchange market, NordPool. With the current good base supply of hydro and nuclear power in Sweden, the deregulation of the Nordic electricity market will probably generate a more stable price level of electricity at the diurnal power exchange market. Smaller price variations between high and low load periods are expected due to the ample supply of regulating hydro power in the Nordic energy system. In other thermal power based systems, as in south Europe, the price differentation will probably be preserved even when the electricity market is deregulated.

As the Nordic electricity market is strongly dependent on hydro power, precipitation will affect the electricity prices for the coming season (1997). A cold winter and spring with very little snow and rain respectively have entailed that the water levels of the water storages are far below their normal levels. The electricity prices for 1997 at the diurnal market are therefore

83 considerably higher than normally. The prices for 1998 are according to the diurnal power exchange market slightly decreasing.

A common belief for increasing electricity prices would make the distributors and the local electricity sales companies invest in new electricity producing units, mainly CHP, since Sweden has a unique demand for both heat and power at the same time. Based on a questionnaire to district heating companies in Sweden, the Swedish District Heating Association forecasts the future expansion of CHP in Sweden. The forecast (1995) tells that the electricity production from CHP will almost redouble until 2010. Due to the taxation system the dominating fuels will be bio fuels and natural gas. Earlier system studies of the Nordic electricity market show that CHP might not be profitable if the deregulation make the prices decrease. Totally it is forecasted that the electricity demand in the Nordic countries will increase with 4 TWh/ year in the nearest future. Together with the coming government decision concerning the phase out of Swedish nuclear power this forecast is an important influence factor on the electricity price.

Model description The linear programming model MODEST, (Model for the Optimisation of Dynamic Energy Systems with Time dependent components and boundary conditions, Backlund 1986, 1988) is originally developed at the University of LinkBping, Sweden. It is a PDC-model (Production, Distribution, Customers), which means that it covers most activities in the energy system from Production via Distribution systems to the Customers. By the modelling, an energy system is seen as a network of nodes and branches. Most nodes represent an equipment like a boiler or a storage and other nodes symbolise a fuel or an energy demand. The branches between the nodes are energy flows of different kinds, in general fuel, heat or electricity. The existing equipment in an energy system and energy supply is compared with a number of potential new installations and energy flows by means of optimisation.

Energy carrier Conversion Distribution Demand

measures

Electricity supply Electricity grid Electricity demand

District hea­ Bio fuel Boiler ting network Heat demand

A schematic view of how an energy system is described by the model

The input data for the model consists of values for fuels and fuel prices, plant capacities and efficiencies, running- and maintenance costs, demand profiles for electricity and heat, investment costs etc. Most parameters can, if needed, be given as

84 time dependent and varied in accordance with a flexible time division, that is chosen in the beginning of the programming. Normally the division below is used .

Long time Short time i j

Season Day No Hour Annual days Annual hours

weekday 1 600-700 98 98 2 700-800 98 98 1 3 800-1600 98 686 4 peak hour 98 98 November 5 1600-2200 98 490 6 peak hour 98 98 to 7 2200-600 98 784

March peak day 8 600-700 5 5 9 700-800 5 5 10 800-1600 5 35 11 peak hour 5 5 12 1600-2200 5 25 13 peak hour 5 5 14 2200-600 5 40

weekend 15 600-2200 48 768 holiday 16 2200-600 48 384

2 weekday 1 600-2200 62 992 2 2200-600 62 496 April, September, weekend 3 600-2200 29 464 October holiday 4 2200-600 29 232

3 weekday 1 600-2200 63 1008 2 2200-600 63 504 May, June, weekend 3 600-2200 29 464 August holiday 4 2200-600 29 232

4 weekday 1 600-2200 22 352 2 2200-600 22 176 July weekend 3 600-2200 9 144 holiday 4 2200-600 9 72

Time division

Because of the peak load during the winter season the time division is more detailed for these months than for the rest of the year. During one time period, i.e. a short time period during a long time period, the state of the energy system is supposed to remain constant concerning energy demand, energy prices and all other properties of the energy system. Which division you choose is though also depending on the electricity pricing and formation of the electricity delivery and pricing contracts.

As a result from the modelling the technically and economically best way to supply the energy system is calculated. Duration graphs, the minimum system cost and the therefrom obtained emissions are calculated. In the modelling existing and new

85 equipment are balanced towards each other, fixed and variable costs considered. The objective of the optimisation is to minimise the system cost which consists of capital costs of new installations and running costs during an arbitrary number of years. Running and capital costs are expressed as their present values and can thus be compared. The optimisation yields the combination of components and energy flows which has the lowest system cost, since this solution is best adapted to the existing equipment as well as to the size and time dependence of energy demand. With a ranging method the optimisation program also indicates within which cost intervals different system solutions are profitable. The method has been applied in the Ronneby analysis.

DSM Measures The DSM measures modelled in this study are load management and efficiency improvements. Efficiency improvements can be described detailedly but are generally described as a possible percentage of the demand profile related to an investment cost (kr/MW). The percentage can be time dependent and related to a factor of influence, which limits the efficiency improvements.

Sector Winter Spring, Autumn. , Summer Industry 1,0 0,9 0,8 Residences 1,0 0,5 0,3 Services and misc 1,0 0,8 0,7

Factor of influence for efficiency improvements of electricity usage in different demand sectors in Ronneby

The influence factors are based on an analysis of the energy system in Vamamo, Sweden. The electricity demand in residences was in Vamamo as well as in Ronneby partly caused by electric heating. The demand in this sector had therefore the strongest seasonal dependence.

Load management is in the model described the same way as storage of heat in an accumulator. Those two activities are the most apparent link between the time periods in the model. Their common denominator is the principle that the sum of all controlled energy flows is equal to zero. As the heat storage can be charged and discharged in any time period certain loads can be cut off as well. Both activities reduce the required output from the production units during some time periods and require increased output during other periods. In contrast to a heat storage though, a storage representing rescheduling of an electrical process can be discharged before it is charged.

The investment cost for load management is given in kr/kWh since the investment cost of making load management possible in many cases is a minor part of the total cost. The major part of the cost is mostly composed of increased labour etc. due to production disturbances.

86 The Energy System of Ronneby The town Ronneby is located 550 km south of Stockholm. The local public utility supplies 250 GWh electricity and 33 GWh district heating a year. The expected maximum demand is 54 MW electricity and 10 MW heat. The electricity is purchased from Sydkraft, (Sydkraft Elforsiiljning AB, the electricity sales company of the Sydkraft family). In the analysis the power and energy charges of electricity are assumed to vary between 0 to + 5%/year, real value. The taxation is basically the one existing in Sweden in May 1996. Both the effects on the energysystem from the existing electricity delivery contract and the forecasted electricity prices of the diurnal market for 1997 is studied in the analysis. In August 1996 the estimated price for the winter season (diurnal market) was 335 SEK. This price is in the analysis only used at peak hours during peak days. All prices are varied with the season and certain time periods (see Model Description). The variations are based on experience from power exchange. The analysis is concentrated on the system benefits of introduction of CHP and DSM measures. One condition in the analysis is that potential DSM measures are not to reduce the elctricity demand with more than 30%.

Generally introduction of CHP requires a heat demand of more than 60 GWh for profitability. The heat demand in Ronneby is therefore increased in the analysis. It is possible to convert required customers to district heating to fulfil this condition. Though in the analysis no regards are taken to eventual conversion costs, however investment costs for distribution systems are included.

This case study has been performed from April 1996 to August 1996. The analysis covers a period of five years ahead. The assumed annual real discount rate is 6%. The aim of the study was to compare the benefits of different investments in the production, distribution and demand sectors by weighing these investment towards each other. About one hundred program runs have been performed to decide the investment frames within different DSM measures are profitable. In the model the energy system of Ronneby is described by ca 2500 different equations. To examine the robustness of the calculated sollutions for the system, the electricity price has also been varied from 0 +5%/year. The effects on DSM and other measures of different strategies for electricity purchasing, for example from diurnal power exchange to long time contracts has also been studied.

To make a more detailed study of DSM measures possible the municipal electricity demand has been devided into the tree sectors; industries, residences and services and miscellaneous. Load management and efficiency improvements were studied in all sectors applying an average factor of influence for each sector and each season.

The potential investments considered were efficiency improvements in industries maximum 7% of their electricity demand, 20000 kr/kW residences maximum 10% of their electricity demand, lOOOOkr/kW services etc.maximum 5% of their electricity demand, 14000kr/kW load management industries maximum 4 MW winter season residences maximum 5,6 MW winter season services etc.maximum 1,5 MW winter season

87 new production capacity efficiency 87% alpha 0,35 investment 25 000 kr/kWe

Results At the current price levels for electricity load management generates a reduction of the system cost of 12-13 MSEK real value in 5 years, investment costs excluded. This corresponds to roughly 3% of the total cost for supplying Ronneby whith heat and electricity. Totally up to 12,5 GWh are mannered, which for the electricity distributor is a clear saving of energy costs of about 700 000 SEK each year. Roughly the same amount of energy is mannered irrespective of incresed electricity prices (up to 5%).

5,6 MW of the residental sector, 4 MW of the industries and 1,5 MW of the service sectors are manouerable in the analysis. Interruptable equipment are in the residental and service sector mainly warm water heater and dual fuel boilers (oil and electricity). The 4 MW in the industrial sector consists of adjustable processes and interruptable electric boilers. The maximum load management occurs a peak day during the winter season at a peak hour between 08.00-16.00 , and between 06.00-08.00 winter weekdays.

Looking at the residental sector, load management also reduces the customers' energy costs. Roughly their share is 223 000 SEK each year distributed among estimatedly 3 300 residental customers. The electricity sales company in Ronneby earns 370 000 SEK/year on buying electricity in low price periods. This means that totally, with the loss of receipts included, the sales company almost earns 150 000 SEK/year only on savings of energy costs. Provided that the interruptable load of the residences is disconnected when the electricity peak demand in Ronneby occurs, so that the peak is reduced, then the total earnings of load mangement of residences in Ronneby would increase up to ca 900 000 SEK/year.

Electricity prices related to the diurnal power exchange market does not favour load management. The small price variations between day, night and weekends and the absence of power related charges reduces the earnings of load management considerably. In the analysis the reduction of the system cost due to load management is less than 1% (4MSEK), the investment costs not considered.

For many customers though, load management will still be very attractive since their electricity delivery contracts contain power related charges.

Efficiency improvements are profitable both with todays delivery contracts and with the electricity prices of the diurnal exchange market. Tire decrease of the system cost is 1, 2-1,6% (5-8 MSEK in 5 years) as the improvements are introduced, the investment costs of 29 MSEK included. Since the electricity prices of the diurnal power exchange market are considerably higher than those of Swedish traditional contracts, it is natural that the economical benefits of efficiency improvments are larger at theese price levels. Totally the electricity demand in the analysis is reduced with a little more than 12 TWh.

The combination of load management and efficiency improvments is naturally most attractive at a large price raise of electricity of existing delivery contracts. The system cost can in theese cases be lowered with around 20 MSEK (4,5%) in 5 years.

88 The most attractive investment alternative in Ronneby is a combined heat and power producing unit. Such a plant also requires the largest investment cost, in this case 100 MSEK for a 4MWe/12MWh unit. With the investment costs included and a write-off period of 25 years, Ronneby would earn about 30 MSEK in 5 years, at existing delivery contracts and no electricity price increase. Related to the electricity prices of the power exchange market Ronneby would decrease their system costs with 36 MSEK/year (5% electricity price raise) if the plant was built.

Together all theese potential investments in load management, efficiency improvments and CHP, (totally 129 MSEK excl load management) could reduce the cost for supplying Ronneby with heat and electricity with up to 57 MSEK, (13,6 %) in 5 years.

References Henning Dag, Energy Systems Optimisation Applied to Local Swedish Utilities, Linkoping Institute of Technology, Sweden, 1994.

Teknikstyming av elanvSndning rapport 1, (Technical control of electricity usage report 1, in Swedish), SEU, Sydkraft 205 09 MalmO, Sweden, 1989.

Prognos 1995, (in Swedish), The Swedish District Heating Association, Stockholm, Sweden 1995.

Curt O BjBrk, Industrial Load Management -Theory, Practice and simulations, LinkQping Institute of Technology, Sweden, 1989

Lennart Backlund, Optimization of Dynamic Energy Systems with Time Dependent Components and Boundary Conditions,

Bjtim G Karlsson, Mats Sdderstrtim, Dag Henning, Simulating av kamkraftsawecklingen pi uppdrag av energikommissionen, Linkoping Institute of Technology, Sweden, 1995

89 Impact of Deregulation - DSM in Industries

Ulrika Bergstrom

Sydkraft Konsult AB and the Linkoping University

90 Impact of Deregulation - DSM in Industries

Ulrika Bergstrom, Sydkraft Konsult AB and the Linkoping University, Sweden

Summary The competition of the deregulated electricity market has increased the suppliers efforts to serve their customers with attractive "added values ” included in the electricity sales contracts. To study the deregulation's impact on DSM measures, computer simulations have been performed of the electricity use and of potential end-use measures at some industries ’ in Sweden. The industrial simulation model INDSIM hasbeen applied on the industries ’ load curves and measures as load management, introduction of dual fuel systems and potential energy efficiency improvements have been simulated. To compare the consequences upon DSM measures from the different market philosophies the analyses are based on both former traditional, monopoly electricity contracts and on typical modern, competitive contracts.

Among others it is shown that the electricity costs for industries hasdecreased considerably after the deregulation and that energy efficiency improvements as a consequence has become less profitable. The same investment brings a lower pay off when the electricity market is competitive. Concerning the competition between oil and electricity for heating, due to price structures, oil has after the deregulation sometimes become more economically benefiting than electricity. This is a surprising result since the average price level for electricity is lower than it was before. Regarding load management, in spite of the lack of power related charges in the electricity sales contracts, these measures can be more profitable after the deregulation. A fact that mainly depends on increased power charge levels in the electricity distribution contracts.

An other consequence of the deregulation is that the electricity sales companies ’ basic motives for DSM measures have become evident; to tie old and new customers tighter to the company through increasing their comfort and reducing their energy costs. The grid owners' incentives for introducing DSM measures are more traditional; to avoid heavy investments due to bottle-necks in the distribution grid or to reduce the own power related subscription costs paid to the regional distribution company. Conflicts may occur within the same group of companies if the sales company activate DSM measures that at the same time effectively reduces the income of the gird owner.

91 Introduction At a competitive electricity market the electricity suppliers try to find attractive "added values ” to offer together with the electricity. These products and services are in a long term ment to tie the customers to the specific supplier. Studies show that industrial customers' main interest concerning their energy supply is safe electricity deliveries which prevents unexpected production disturbances. Different products and services are therefore offered from the suppliers to fulfil that request. Secondary, these customers want to reduce their energy costs and the suppliers give advice and offer services of increased energy efficiency and improved energy management. Generally, the hardening competition of the deregulated electricity market has increased the suppliers efforts to serve their customers.

To study the deregulation ’s’ impact on DSM measures and as a link in EnerSearch’s aim in trying "to meet customers' desires through IT solutions" computer simulations have been performed of the electricity use and potential end-use measures at some industries' in Sweden. The analyses are made in co-operation with the local distribution utility and both traditional electricity contracts of the monopoly market and new typical deregulated contracts are simulated. Load priority systems, dual fuel systems and potential efficiency improvements are examples of measures included in the analyses. A general question to be answered is how the deregulation already have and in the future will influence DSM co-operation.

Methodology The industrial load management simulation model INDSIM hasbeen applied for the analyses. INDSIM is developed at the department "Energy Systems” at the Technical University of Linkoping, Sweden, and is mainly applied for studies of effects from industrial load management. It is applied at one company at a time and can among others simulate load priority systems, dual fuel systems, short time accumulation, load rescheduling, efficiency improvements and internal combined heat and power production.

INDSIM

Load Data End-use Electricity measures rate

Creation of Simulation of Electricityuse load curve and end-use measures

New Energy System costs for load curve different combinations of end-use measures

The simulation procedure of INDSIM

92 As input for the analyses the company’s annual demand profile (8760 load values) and electricity contract are specified. The annual load curve is mostly constructed outgoing from one or two weeks hourly measurement results. The data are expanded to a one year load data base, either through a random process or through assuming that the recorded load curve is constant all weeks of the year. In the random routine consideration is taken into the ratio between the recorded one hour maximum load and the yearly maximum power demand when the load curve is calculated. After the first estimations of the annual electricity demand it is also possible to adjust the load curve so that it corresponds with the actual demand. The load data are completed with facts about industry specificequipment and measurement test resultswhich are collected at one or several company visits. At the surveys possible energy saving potentials are mapped out together with data of adjustable equipment etc. These end-use measures can in the model be associated with their respective investment and maintenance costs as well as with their expected economic life. When the required industrial surveys are performed and the simulation strategy outlined is INDSIM applied on the company's electricity demand profile.

Simulation strategies The industries are analysed in order to

1) Discover economically benefiting DSM measures 2) Examine the DSM measures’ dependence upon the electricity contracts ’ different price structures.

Which measures that are included in the analyses is depending on the conditions of the specific industry. Mainly efficiency improvements, load management and introduction of dual fuel are the most interesting alternatives. The following electricity contracts are analysed:

• Traditional electricity contract before the deregulation; sales and grid charges included • Typical industrial electricity contract (1997) after the deregulation (only sales charges and no power related or subscription charges)

» Electricity grid contract after the deregulation

• Combined sales and distribution contract after the deregulation

Power Demand Profiles - DSM options Company 1 The first company analysed has a relatively smooth power demand profile since the production is running almost all year round. During a week the power demand normally keeps around 1600-1700 kW +/- 200 kW. An average month the company consumes a little more than 1200 MWh. An exception is the vacation month which only hasa 15% electricity demand compared to the normal level. During the year the electricity demand is about 13,8 GWh. The load is climate dependent and the maximum demand is 2286 kW (1997).

93 Load profile

Load profile one week in January consistingthe extreme

power peak of the year.

Load management and efficiency improvements are the most interesting end use measures at the company. Manageable objects to start with are electrical boilers and heat pumps, since this equipment mainly serve thermally slow systems (heating of the premises). Other interesting objects to manage are process related equipment like blast engines and converters. Normally the converters are run most of the time at partial load which leaves many opportunities for load rescheduling. The capacities of the electrical boilers are 2 x 300 kW and 2 x 350 kW. Provided that process related equipment are taken into account as well the company’s maximum power demand can be reduced with approximately 800 kW for 1-2 hours periods. The structure of the electricity contract constitutes the pressing force if this extra equipment will get included in the load management system or not.

Major savings can also be achieved through improved energy efficiency. The concerned items are mainly heat recovery from cooling systems and air compressors, improved dimensioning of air compressors, alternatives to compressed air for PVC-strips transportation and regulation of heat and cooling systems for cylinder rolls in calander lines. Approximately can energy savings up to 30% be performed.

Simulations are performed to examine the economical benefits from load management of different extent and efficiency improvements.

Company 2 In opposite to the first company, the second company's demand profile has large variations caused by process related vacuum furnace cycles. During the day and night the power demand varies +/- 700 kW. The energy demand is even all year round about 540 MWh/month, 83 % in July (vacation). Totally the company consumes 6400 MWh with the maximum power peak 1590 kW.

Potential DSM measures are load management, introduction of dual fuel system and heat recovery. The company uses 3 vacuum furnaces, each at 680 kW in the production and the simulations will show the economical benefits from load rescheduling of this equipment. Only for heating of premises the company consumesabout 240 m3 oil, although there are 3 electrical boilers at 560 kW each. Due to Swedish taxation system oil for industries is cheap

94 (1500 SEK/m3) why the electrical boilers are not competitive part of the year. Simulations are performed to examine if the electrical boilers can become a complement to the oil based boilers. Concerning heat recovery the surplus heat from the vacuum furnaces is blown off in a cooling tower at the roof of the building. Approximately the loss of energy corresponds to 2352 MWh. If the heat instead was recovered or sold to the energy utility, the company would get a small income.

Load Profile

Load profile one week in January

Since the dual fuel system and the heat recovery system are different alternatives to gain benefit from corresponding systems in the factorythese alternatives are not possible to run at the same time.

Analyses Results

Company 1 With no measures activated the company's simulated electricity demand 1997 is 13,8 MWh and the maximum power peak during the year 2286 kW. The utilisation time is 5973 hours and the load factor 68%. Before the deregulation the total cost for electricity was 4068 kSEK (demand costs 798,2 kSEK, energy costs 3035,6 kSEK subscription charge 235 kSEK) with an average cost of 298 SEK/MWh. Based on the new sales and distribution contracts after the deregulation the total cost is almost 2,5% lower (3970 kSEK) and the average cost 2,35% (291 SEK/MWh) lower than before.

When a load management system is activated and the load kept under 2050 kW during the critical seasons, the company saves about 32 kSEK/year (0,8%) with the former contract and 59 kSEK (1,5%) with the new agreements, mainly due to the cut peaks. The peaks are at the most reduced with almost 250 kW. But a cold winter day it is probably possible to cut the maximum power demand even more provided that process related equipment is taken into account. With a 500-600 kW reduction the company would save about 200 kSEK/year - former contract - and 235 kSEK/year with the new agreements.

95 The increased savings are due to new higherpower related charges from the still monopoly distribution utility and former high fixed charges that earlier blocked the benefits from load management. The lower subscription charges do not effect this result. But, because of the decreased energy prices load management becomes a little less profitable compared to earlier when not only the peaks are cut but energy is managed as well.

The figure shows that load management is most profitable with the new

contracts and when "only" power and not energy is managed.

A classical DSM measure at the monopoly market was to change the electricity contract between the company and the energy utility so that the incitements for the company to activate load priority systems were extra large at times when it gained the energy utility. Applied in this case an appropriate measure would have been to adjust the power charge of the former contract so that it was only charged between 8-10 a m and 16-18 p m during the winter season. Provided that such an agreement was possible and that the power peaks do not exceeded 1500 kW at these periods, the company would have earned almost 200 kSEK in a year. For the utility the co-operation would have ment savings about 40 kSEK (240 kSEK from saved power costs minus 200 kSEK from lost income). Approximately the power peaks would be reduced with 700kW winter months and 200 kW reduction during the summer.

At the deregulated electricity market these kind of DSM measures are less interesting for the electricity sales companies since there are no power related charges at the electricity trade market. Many of the local energy utilities and industries also have electricity contracts that in some way are related to the price structure and price trend at the trade market. On the other hand it is still an interesting alternative for the distributor who has power related subscription costs that he pays to the regional grid owner. It is also an especially important alternative to heavy investments if the concerned power lines are under-sized, a situation though not very common in Swedish distribution grids.

96 24 hours load management 8-10,16-18

hour

24 hours load management down to 1500 kW between 8-10,16-18, original and adjusted load curves

Effects upon the energy utility from load management between 8-10,16-18

50000 48000 46000 44000 42000 g 40000 38000 36000 34000 32000 30000 t— C0L0h-0)T— COlOh-Oi— CO r— t— t— CN CN hour

The utility's original load curve and reduction due to load management

Finally efficiency improvements would continuously save the company a lot of money. A 10% reduction of the energy demand ment 407 kSEK/year savings in energy costs with the former contract With the new structure the company would earn 397 kSEK/year. Approximately it would be possible to reduce the demand up to 30% which corresponds to 1220 kSEK - former contract - and 1191 kSEK - new contract, investments excluded. A consequence of the deregulation is that the lower electricity prices for industries also make energy savings for

97 thesecustomers a little less economically benefiting. The same investment costs generated a better pay off before the deregulation. This might lead to that investment money are focused to other - not energy related - areas in the company.

Company 2 With no measures activated the simulated electricity demand is 6,8 MWh and the maximum power peak 1585 kW (1997). The utilisation time is 4266 hours and the load factor 49%. Except the electricity costs the company pays about 360 kSEK/year for oil to heat the premises. The total cost for electric energy was before the deregulation 2316 kSEK (demand costs 636 kSEK, energy costs 1668 kSEK, subscription charge 12 kSEK, VAT excluded) and the average cost was 343 SEK/MWh. With the new contracts the company immediately saves almost 3,5% (80,5 kSEK) of their former costs and the average electricity cost is 331 SEK/MWh.

When load management of the furnaces is activated and the load is kept under 1200 kW the peaks are reduced with ca 300-350 kW and the company saves about 147 kSEK/year (6,4%)with the old electricity contract. With the new agreements the company only saves 1 kSEK more than before, but 6,6% of the total electricity costs. The grid related power charges are suited to fit the old level of the electricity contract. This means that the power related charges have about the same part of the total electricity costs as before (27%). A consequence is that the profitability of load management is almost equal. Due to a relatively small fixed charge load management has become more benefiting than earlier. But the higher electricity prices of the monopoly contracts compensate this profit as soon as not only the peaks are cut but as well energy managed.

Introduction of dual fuel

•------original load curve ------adjusted load curve

-t " !' i. Ji,'i ?ii I " ti tin ..... I " 'i -y „

timmar

Company 2 with original and adjusted load curves

Since the company already had a very attractive agreement with the energy utility it was of no use to change the electricity contract so that the power demand charge was active only when the utility had power peaks. The measure would only bring that the company used more energy (43 MWh) at other time periods which would reduce their earnings due to the load priority system, 135 kSEK compared to 148 kSEK before. For the utility the small peak reduction would reduce their costs with ca 140 kSEK. Their loss of income is 135 kSEK, why the total gain

98 only becomes 5 kSEK/year. The incitements for the distributor were equally small since the concerned power lines were well dimensioned.

A better alternative for the company would be to utilise the existing dual fuel system and to run the electrical boilers within the power limits of the electricity contract Before the deregulation the savings from such a measure would have been better than after the deregulation due to former higher price differences between different time periods and generally lower bottom price levels. Even though the company would have used ca 360 MWh more electricity, their total energy costs would have been reduced with almost 110 kSEK (4,1%) and the electrical boilers had replaced approximately52 m3 oil. The running time for the boilers would be 3880 hours. With the new agreements the electricity is cheaper in general why it is easyto believe that the oil alternative would be less competitive. But, since the bottom price level during winter was lower before the deregulation and the price differences s between certain tariff periods hasbecome smaller the electrical boilers now only replaces 32 m3 oil, which brings savings of approximately 100 kSEK/year (3,9%).

Although if both the load priority system and the dual fuel system were introduced the total cost reduction for the company would be almost 250 kSEK/year independent of which contract that is simulated (9,3% old contract and 9,6% new contracts). The same investment brings a better pay-off at the deregulated market The oil replacement though would be reduced to 21 m3 with the old contract and stay the same (32 m3) with the new contracts.

The best option for the company is to sell or recover the surplus heat from the vacuum furnaces. Provided that the heat is sold to the energy utility it would have a value of ca 470 kSEK. Here it is assumed that the heat during the summer months replaces oil based heat in the district heating system, and in winter replaces the heat from a bio fuelled hot water plant

Comments Since the electricity costs for industries have decreased after the deregulation the benefits from efficiency improvements are reduced as well. The logical conclusion would be that investment capital within companies is focused to other- not energy - related areas due to the longer pay -off times. A fact that contradicts this development in Sweden is the generally higher electricity price levels in the rest of Europe. Theselevels will be "imported" to Sweden as soon as the European electricity markets are deregulated as well, and there are no bottle necks in the transmission that can influence Swedish electricity export.

Regarding load management several studies have shown that reduction of power peaks would become less profitable at the deregulated market due to the lack of power related charges in the electricity sales contracts. Theoretically this is true, but practically many of the distributors have kept or increased the level of the power charges in that extent that load management has become even more profitable for the companies than it was before. Also concerning the competition between oil and electricity for heating it is a bit surprising to find that even though the average price level for electricity is lower after the deregulation, the price structure is such that it during a year sometimes has become more profitable to run the oil based boilers than the electric alternatives.

99 In the future it is not very likely that severe conflicts concerning DSM measures/load management occur within a energy combine of companies or between the sales and distribution companies that have contracts with the same customer. But it is important to be aware that the reasons for the sales company to introduce these measures are totally different from the reasons that the grid owner might have. The sales company sees DSM as a step in their work to reduce the customers energy costs and to tie the customer tighter to the company. A consequence is that the cost reduction mainly will be taken from the income of the grid

owner that now is an independent company with it's own result responsibility. The grid owner's main reason

to introduce DSM and reschedule load is if there is a bottle neck somewhere in the distribution grid or in traditional DSM-manner to reduce own power costs by changing the customers ’ load curves.

Conclusion The electricity costs of the simulated industries have after the deregulation decreased up to 3,5% compared to the costs at the monopoly market. As a consequence the profitability of efficiency improvements has decreased as well. The reduction is not very big - up to 2,5% - but it brings longer pay off times for efficiency improvements which might lead to that the limited amount of investment money within a company get focused to other - not energy related - areas.

Although the average price level of electricity is lower than before the deregulation, electricity has part of the year become less competitive to oil. At the monopoly market electrical boilers could be run a lot more and with better benefit during a year mainly due to lower electricity prices during spring and summer. In the winter season, the prices have been reduced considerably after the deregulation, but then the customers instead reach the power level of the distribution contract which of course limits their willingness to consume more electricity. The increased demand in winter can not compensate the decrease due to higher prices the rest of the year. Regarding dual fuel systems, oil as a complement to electricity hasduring winter - when it can be used as a tool to control power peaks - become less attractive to the customers. In spring and summer, it is more competitive.

Concerning load management, in spite of the deregulated market’s lack of power charges in many industrial electricity sales contracts, it is shown that these measures after the deregulation can become even more economically benefiting for the customers than before. The increased profitability is due to sometimes higher power related grid charges and a loss of former large fixed charges in the sales contracts. These fixed charges had a buffering effect that negatively influenced the profits from load management (%).

It is also shown that the profitability (in percent) of industrial load management after the deregulation mainly increases compared to earlier as long as power peaks are cut but not energy managed. Thisdepends on the in average lower electricity prices after the deregulation. If large amounts of energy were managed as well, the former higher electricity prices would compensate the new power related grid charges and make load management before and after the deregulation equally profitable.

Regarding load management based co-operation between electricity utilities and industrial customers to reduce joint power peaks, after the deregulation the distributor has become the new partner for the customers. As the sales companies no longer pay any power charge, in short term, load management only brings a reduction of income for them. Instead, their basic motive for DSM measures is to tie old and new customers tighter to the

too company. For the sales company all cost efficient measures that reduce the customers ’ electricity demand are important tools - added values - to be found in the electricity contracts portfolio. The reasons for the grid owner to Introduce DSM are mainly due to bottle necks in the distribution grid or to reduce the power related subscription costs paid to the regional distribution company.

The conclusions are drawn from the results of the demonstrated cases and based upon experiences from earlier simulations.

References D. Suleyman, "The impact of a deregulated European electricity market on Volvo in Sweden", LinkOping University, Paper at Intersociety Engineering Conference on Energy Conversion, Colorado Springs 1998

B. G. Karlsson, M. Soderstrom, D. Henning, "Simulering av karnkraftsavvecklingen pa uppdrag av energikommisionen", LiTH-IKP-R-888, December 1995

D. Henning, Modest an Energy-System optimisation model applicable to local utilities and countries, Energy-The international Journal, Vol. 22, No. 12, pp. 1135-1150,1997

U Bergstrdm, "Simulation of a local energy system with focus on cost-efficient DSM measures on a deregulated electricity market", SydkraftKonsult AB, DA/DSM Europe 1996 Volume 3

D. Henning, Cost minimization for a local utility through chp, heat storage and load management, International Journal of Energy Research, Vol. 22, No. 8, pp. 691-713,1998

M Andersson, "Cost-effective incentives for local electric utilities and industries in co-operation, Modelling of technical measures, LikOping University, 1993:12

K Bjdrk, "Industrial load management simulation", LinkOping

101 Traditional load management versus real time load control

by

Ulrika Bergstrom

Sycon Energikonsult AB and

the Linkoping University

102 Summary The rapid development within information technology enables a more dynamic and flexible operation of energy systems. Therefore thequestion has arisen of how the energy system in it’s entiretywould be affected by instant price variations and load management.

In order to examine the effects from real time load control and real time electricity pricing versus traditional load control and tariffs, computer simulations have been performed of load management measures in the energy system of Ronneby, Sweden. The analyses are made outgoing from two synthetic load curves and coincident prices based on actual demand and spot market electricity prices in December and May 1997. In the performance the energy system optimisation model MODEST has been applied on the energy system.The extent of potential end-use measures are estimated by the local distribution utility Ronneby Energy AB.

It is among others shown that power related charges constitute the determinant factor for how and when load control shall be activated. The reduced energy systemcost due to load management is therefore larger with the old electricity price structure than withreal time energy pricing from the power exchange market.

During the part of the year when the power related charges are not valid customer driven load control against real time spot prices might generate unexpected distributor load profiles. Bottle necks in the distribution grid that can be avoided the part of the year when the management is distributor driven, might therefore cause new problems in new time periods.

If real time pricing was introduced to customers with controllable equipment, thesecustomers would utilise their load priority systems the whole year round and not only during the cold season. The benefits from real time pricing corresponding to real time load management is totally dependent of the variation of price and not on seasonal movements or the electricity price level.

103 Content

Introduction

Description of the Ronneby Energy System

Price structures

Methodology

Analyses

Result

Conclusion

104 Introduction As the electricity market became deregulated and the distribution and the sales of electricitywere split into separate companies also new electricityprice models were introduced to local electricity sales utilities and to industrial customers. To begin with, the distribution and sales related charges were separated in the customer tariffs, a division that had caused many discussion at the time of the company separation. Then, due to the competition, the structures of the sales agreements started to change while the monopoly distribution charges remained more or less the same. Some of the new developed price models became connected to the trend at the power exchange market. A next step might be to make a large scale introduction of real time pricing. As the rapid development within information technology (IT) also outgoing from a cost perspective will enable real time load control and power generation the question has arisen of how a more dynamic electricity pricing would influence the energy systemin it’s entirety.

The following report is focused on load management within the concession area of a local distribution utility. It’s main object is to view the impact on the distributor from load control against traditional tariffs versus real time electricity pricing.

Description of the Ronneby Energy System The cityof Ronneby is located in south Sweden. Theenergy system consists of an electrical consumption of approximately 260 GWh electricityand 35 GWhheat. In 1997 the local public distribution utility, Ronneby Energy AB, experienced a maximum power demand of 52,3 MW.

Duration Graph Ronneby Energy AB 1997

60000

50000

40000

5 30000

20000

10000

CN CO lO (O

As this case study is concentrated on electrical load control, the Ronneby district heating system will be given no further consiedration.

To make a more detailed study of load management measures the municipal electricity demand has by the utility been divided into four customer categories; industries, residences, services and miscellaneous and street lighting. The aim with the division is to obtain a more flexible system where the effects from load control within each customer group can be analysed. A finer time division during 24 hours is

105 thereafter approximately performed. Load management have been simulated in the three first sectors by applying possible maximum peak cuts and probable limits for load rescheduling.

OPTIMISED ENERGY SYSTEM

load residences

buffer! PRICE MODEL residences load control node

Spot price

Traditional tariff buffer! industries load control Distribution node Distribution price

load Power related charges utilities

buffer! Services load control node

load street lighting

Manageable object in the residential sector are electric water heatersand dual fuel boilers (oil and electricity). The first have a maximum output of 3 kW. Based upon experiences they are expected to be run at 1,5 kW in average. Considering the simultaneity factor it is estimated that around 1 kW/ villa can be managed. The assumption is made outgoing from Sydkraft experiences from similar projects in Staffanstorp and Vamamo. In Ronneby there are around 3300 residences with potentially controllable water heaters. This leaves a potential of 2,5 MW load management provided that only heat exchangers are controlled.

Around 1300 customers have combined boilers. Their maximum power outage is 300 kW, but are expected to have an average controllable power of 150 kW. Provided that 1-2 % of these customers utilise electricity during the winter season, still around 3,1 MW more load can be controlled within the residential sector. Totally 5,6 MW can be managed all year round and approximately around 2,5 MW during the summer.

Controllable objects in industry mainly consists of adjustable processes and interruptible electric boilers. REAB estimates that around 4 MW can be controlled during the winter. In the service sector around 1,5 MW in water heaters and combined boilers can be controlled.

Price Structures Activation of load management depends on the price structure. Two main models are applied in the analysis; the traditional tariff and the spot price of the power exchange market.

106 Traditional Swedish tariffs are divided into 3 or 4 seasons, with one to two price levels during the day and night. During winter, spring and autumn there are one "high-price period” between 06-22 and one low-price period between 22-06. At weekends and holidaysthe low-level is charged. The distribution charges were until the deregulation included in the tariff but are since January 1996 billed separately. Except distribution charges the tariff also contained power related charges in forms of fixed charges, subscription charges and power charges, depending on the customer category and size of themain fuse.

Below a typical tariff with the characteristic timedivision.

Time Period Energy Price

November - March, weekdays, 06-22 280

November - March, remaining time 223

April, September, October, weekdays, 06-22 223

April, September, October, remaining time 161

May- August 161

Power charge, 6 hour value November - March, 06-22: 260 000 kr/MW Subscription charge: 15 000 kr/MW Fixed charge:

Due to the deregulation electricity sales contracts connected to the price trend at the power exchange have been realised both for distributors and industrial customers. The physical exchange atNordPool is divided into Day-Exchange (the spot market) and Regulate-Exchange. At the spot market hourly electricitycontracts are traded for the following day-and-nights ’ 24 hours. The Exchange therefore result in 24 different prices for every hour during the next day and night.

These prices are based on hourly supply curves that are delivered to NordPool from the balance- responsible companies the daybefore the actual trade. At NordPool hourly demand curves are calculated whichtogether with the supply curves forms a price cross - thespot price.

107 DA YEXCHANGE PRICE CROSS MODEL

SPOTPRICE

HOURLY SUPPLY CURVE FROM THE COMPANY

PRICE

FINAL DEMAND CURVE FROM NORDPOOL

POWER

The Regulate-Exchange corresponds to the price-area and is characterised by a considerably lower turnover (about 1% of the spot market.) In opposite to the spot market it handles the necessary adjustments required when the situation does not follow the supply and demand plan. Such examples are if a plant suddenly falls off or at sudden power transfer limitations. The regulate exchange is an hour to hour exchange and relatively big price differences can occur from one hour to the other. Sometimes producers try to regulate their production units against the situation at the Regulate-Exchange market. For example in that aim Sydkraft sometimes controls certain processes at theback pressure plant Karskarsverket.

As the aim with the report is to in theory reflect how load management ’’normally” would be utilised against real time pricing compared to traditional tariffs the electricity prices of the spot market have been chosen to visualize thereal time price mainly due to the spot markets increasing and considerably higher turn over compared to the regulate exchange.

Average System price variations spot market 1996 - 1997

lO CD O ID O CO

The picture shows the price tendency at the spot market 1996 to 1997.

108 Methodology To view the effects upon load management from different electricity price structures, one day and night have been chosen for in detail examine how load optimally can be controlled from hour to hour. Since the daily load profile varies during the year the aim is not to mirror a day and night representative for the whole year. Instead one ’’normal ” winter and early summer day are chosen to study both the influences of power related charges as from energy price variations. Each dayis though representative for the belonging month.

Also the electricity price at thepower exchange market varies. To obtain representative prices and load curves, synthetic curves have been created outgoing from actual loads and prices during the two months weekdays.

First an average 24 hour duration graph and an average 24 hour load curve are created outgoing from actual hourly measurements. The basic principle is that theload values are taken from the load duration curve and the hour when the values occur are taken from themonthly average curve. This means that the method reflects both the actual value and when the demand occur.

The methodology is originally developed by Casper Kofod at Denmark’s Technical University and further developed within Sydkraft. To achieve an appropriate correlation between the power exchange spot prices and the utility load the spot prices have been sorted together withthe load values to form a representative, synthetic price curve as well - a coincident price.

After the calculations of synthetic load curves and coincident prices the energy system optimisation model MODEST is applied separately on each24 hour period. MODEST is based on linear programming and makes it through various variables and a relatively free node creation possible to give an almost perfect overview of any energy system. Load curves, prices, investment costs, fixed and variable costs as well as technical capacities of existing and potential equipment serve as input. The optimisation yields the combination of components and energy flows whichhas the lowest system cost, since this solution is best adapted to the existing equipment as well as to the size and time dependence of the energy demand.

In this case focus is laid upon load management which in the model is described as a potential electricity storage. This means that an energy balance defines the ’’storage content” as well as charge and discharge during each time period. Eventual losses are considered as well as maximum input and output. Thereby, the function can reschedule the output from the electricitygeneration plants or from the electricity supplier to other periods, with lower electricity prices. As output the model delivers a system cost, duration graphs of loads and how certain units should be run to obtain the minimised system cost.

The results are thereafter analysed and eventual new simulations performed.

109 Analyses The 1997 electricity load of Ronneby Energy AB has been simulated in order to examine the benefits from load management due to different price structures. The analyses covers a period of 24 hours and the case study has been performed in November 1998.

To examine the optimal operation of load management according to the price models two synthetic days and nights have been created outgoing from measured values of two months; December and May. December has been chosen, as being one of the most climate depending months of the year, why power related charges will be significant for activation of load priority systems. The spot price level of December is also less influenced than the January level of thewater lack in 1996. May has been selected as being a late spring month, less affected of distribution charges and can be seen as representative also for months in early autumn.

Based on actual load data of May and December 1997 and spot prices for the same periods synthetic and average load curve have been calculated, see Methodology. The good coincidence between the curves indicate that the synthetic curves well represent the ’’normal ” load, load profile and prices for the actual months.

Synthetic Daily Load Curve and Average Daily Load Curve Week Days in May and December 1997

50 000 45 000

40 000 35 000 - - - Average Load 30 000 Synthetic Load Curve 25 000 - - - Average Load 20000 Synthetic Load

110 Synthetic Daily Price Curve and Average Daily Price Curve Week Days in May and December 1997

- - - Average Price H 150 Coincident Price - - - Average Price Coincident price

The synthetic load curves and coincident prices are utilised as input to theoptimisation model. Load management has been analysed corresponding to five different price structures;

• Spot prices, distribution prices excluded • Spot prices, distribution prices included • Distribution tariffs • Traditional tariffs • Traditional tariffs, distribution prices excluded

Results December For December, it is shown that distribution related power charges are of great importance for the activation of load management. Looking at the effects of load control against the spot price, a load curve that might cause the distributors a lot of trouble and investment costs in extended transfer capacity, is the result.

m Spot price related load management

60000

50000

40000

5 30000

20000

10000

- - - Original load curve ———Manipulated load curve

As low electricityprices occur at times when the distributor has power peaks, the customers will reschedule their load to these periods of large disadvantages for the grid owner.

Coincident Spot prices

hour

Looking at the duration graph the problem of the distributor becomes even more evident. His peak demand increases with 6 MW a ’’normal ” winter day. The same increase at his maximum peak day could become very expensive.

112 Duration Graph, spot price related load management

60000

50000

40000

5 30000

20000

10000 --

Original load curve •Manipulated load curve'

But during thewinter months the customers behaviours as well as the distributors are regulated with the power related subscription charges of the distribution contracts. As these charges are added to the spot price, the load curve get a totally different appearance.

Load managemet towards spot and distribution prices

50000 45000 40000 35000 30000 | 25000 20000 15000 10000 5000 0 roioh-oix — comh-0)T— co •*- T— T— t— t— cm CN hour - - - Original load curve • Manipulated profile

Here the peaks instead are cut with around 6 MW, which result in a smooth energy profile, well adjusted to avoid large variations in electricity generation operation.

113 Duration Graph, load management towards spot and distribution prices

50000 45000 40000 35000 30000 | 25000 20000 15000 10000 5000 0 ■<— COLOh-Ot— COLOh-Ov-CO r r r r r- OJ (vj - - Original load curve .... —' Manipulated load curve ;

Optimising theoperation against the traditional monopoly tariffs the results are similar as long as the power charges are due. It is therefore evident that the distribution tariff alone is the determinant factor for how the load should be controlled.

Load management forwards traditional electricity contracts

50000 45000 40000 35000 30000 | 25000 20000 15000 10000 5000 0 *< — coioh»o)v-coior'-o)v-co T— X— T— T— t— CN CN hour - - - Original load curve ■— 1 Manipulated load curve*

The same result is seen when only the distribution tariffs serve as boundary limit.

114 Load management towards distribution tariffs

50000 45000 . 40000 . 35000 30000 .. 25000 - 20000 15000 .. 10000 __ 5000

- - Original load curve Mainpulated profile

Neither the energy prices of the traditional tariffs generate a load profile desirable for the distributors. Even though theseprices only varies between day and night from a high to a price level more than 20% lower.

Load management towards traditional tariffs power related charges excluded

50000 ,------45000 -- 40000 . 35000 .r** 30000 | 25000 .. 20000 15000 -

10000 -- 5000 ..

- - - Original load curve ■ Manipulated load curve (

May Looking at May it is assumed that the power related subscription charges of the distribution contracts are no longer valid. The higher of these charges is only charged between November to March and based on the maximum one hour peak value during this time. Therefore the distributor and the customers are free to control their load as they prefer, as long as they do not exceed the annual power subscription, which would lead to large punishing fees.

115 In May the highest electricity prices at the spot market coincide with the high load hours. The maximum demand that occurs between 7-13 is therefore reduced with 2,5 MW - the maximum controllable load.

Spot price related load management May 1997 40000 r- 35000 ■ 30000 ■ 25000 ^ 5 20000 15000 -

10000 - 5000 -

hour Original load curve ■ ■ • Manipulated load curve

24 hour synthetic prices May 1997

160

120 .-

Even though sudden peaks are generated later in the day, the duration graph looses some of its’ inclination.

116 Duration Graph, load management towards spot prices May 97

40000 35000 30000 25000 5 20000 15000 10000 5000 0

t* r r r r N OJ hours —Original load curve - - - Manipulated load curve

If the load is rescheduled against the traditional tariff, nothing will happen since the energy prices during May-August are even.

System costs Studying the earnings from load management during these 24 hour periods, it is evident that it is the reduction of the power related costs that causes the largest incomes from load control. As the traditional agreement between the distributor and the electricityproducer contained higher power related charges, load management is about twice as profitable operated against these tariffs.

In estimating the savings during one year it is assumed that the year consists of 1 peak day when the power related charges are activated, 149 normal December days- without power related charges - and 215 dayssimilar to May also without power related charges. The distributor would earn around 2 MSEK from load control against traditional tariffs compared to 1 MSEK if he controlled the load against spot and distribution prices, investment costs excluded.

Looking at how and when the load control is activated the main difference occur during the summer month when traditional tariffs offers a constant price but the prices of the power exchange continuously varies through out the year. In the first case no load will be rescheduled at all, since there are no economical incitements to do so. It is hereby assumed that the power charges are already activated due to high load periods during the cold winter season.

An other difference is that due to the variations of the spot price, the maximum amount of load will be controlled at the times of the highest electricity prices - provided that the power related charges are withdrawn. As the only price difference in the traditional tariffs occur between 22-23 in the evening and 5-6 in the morning, the load reduction will occur during the day but will more have the profile of an efficiency improvement as it is the same how the load is controlled as long as the same amount of energy is managed.

117 Conclusion Power related charges are the determinant factor for load control. As such charges are taken away from the electricity sales contracts, the potential income from load management is generated due to the structure of the distribution contact. The instant gain for a distributor that activate load management will thereby come from the reduction of his own power related charges to the regional grid owner. An other main reason to perform load control is of course to avoid or postpone heavy investment in extended distribution capacity.

Customer driven load control against real time spot prices might generate unexpected distributor load profiles, during the part of theyear that the power related charges of the electricity distribution contracts are not valid. Bottle necks in the distribution grid that can be avoided the part of the year as the management is distributor driven, might therefore cause new problems in new time periods due to simultaneity factors of electricity load.

If real time pricing was introduced to customers with controllable equipment, the load priority systems would be utilised during the whole year and not only during the cold season. Thebenefits from real time pricing corresponding to real time load management is totally dependent of the variation of price and not on seasonal movements or the electricity price level.

References Kofod C, Opbygning av syntetiske belastningskurver, Danmarks Tekniske Hojskole, Lyngby, 1982

Andersson M, ’’Cost-effective incentives for local electric utilities and industries in co-operation, Modelling of technical measures, Likoping University, 1993:12

Henning Dag, Cost minimization for a local utility through chp, heat storage and load management, International Journal of Energy Research, Vol. 22, No. 8, pp. 691-713, 1998

Teknikstyming av elanvandning rapport 1, (Technical control of electricity usage report 1, in Swedish), SEU, Sydkraft 205 09 Malmd, Sweden, 1989.

Bergstrom Ulrika, Impact of deregulation - DSM in Industries, Sydkraft Konsult AB, DA/DSM Europe 1998

Henning Dag, Modest an Energy-Systemoptimisation model applicable to local utilities and countries, Energy-The international Journal, Vol. 22, No. 12, pp. 1135-1150,1997

U Bergstrom, ’’Simulation of a local energy system withfocus on cost-efficient DSM measures on a deregulated electricity market”, Sydkraft Konsult AB, DA/DSM Europe 1996 Volume 3

118