Outages in the electric grid can be costly for society. Because of this, reliability is one of the parameters used to regulate network companies. Reliability in electrical grids can be measured with SAIDI, system average interruption duration index, which for Fortum Distribution has increased in Stockholm’s distribution network (11 kV) in the early 2000’s. By using outage data for Stockholm from 2011-2013, sets of parameters to be used for reliability calculations in the network information program Tekla NIS were derived. Two different options for investment were then analysed: changing old cables with high fault frequencies, and installing automatic switching in distribution substations. These options were also analysed in a simplified model built in Microsoft Excel. The model calculates the reliability of Stockholm’s distribution network using Markov chains and the network’s average line from a transmission substation’s feeder. The Excel model is faster than Tekla NIS but is limited and less detailed, however the results from Tekla NIS and the Excel model were found to be almost equivalent. The priority for reliability investments in Stockholm should be to change old cables of the type FCJJ since it increases the fault frequencies in the network, while an ageing network decreases the grid fees Fortum Distribution can charge. Simplifications and assumptions, due to insufficient outage data and some problems with Tekla NIS, had to be made when deriving parameters for reliability calculations in both Tekla NIS and the Excel model. However, Tekla NIS can be used to incorporate reliability analysis in network planning, but since Fortum Distribution currently only uses it sparsely, further testing and analysis of the programme is recommended if the usage is to increase. The Excel model can be used for making fast and rough estimates of the result of the two analysed options of investment. / <p>I den tillgängliga fulltexten är tre figurer med tillhörande information i kapitel 2.4 borttagna från originalrapporten efter önskemål från samarbetspartner.</p>

The distribution of electricity is a natural monopoly. The infrastructure of the electricity grid is divided into areas and the distribution system operator (DSO) has concession for the distribution of the defined area. The concession is governed by laws and regulation.The Swedish Energy Markets Inspectorate (Ei) is the national regulatory authority. The continuity of supply of electricity is a part of the regulation. The incentive is given by a revenue cap regulation which may result in a reward or penalty.The performance indicators are a measure of the continuity of supply and used in the calculation of the revenue cap regulation.The Master thesis aims to specify the possible consequences for the electricity distribution system operators and their customers utilizing the electricity grid in case of a change of the current methods used to measure the continuity of supply in Sweden.With the current methods, the Ei regulation for year 2016-2019, is the continuity of supply in the local electrical grid estimated from a view where each disruption is treated equal and therefore is SAIFI and SAIDI used. In the regional electrical grid is another method used to measure the continuity of supply, estimated from a view where the loss of supplied energy is considered, therefore uses PNS and ENS. In the documentation is the term ILE used for ENS and ILEffekt for PNS.The Master thesis is constructed to analyze new performance indicators for the local and regional distribution grid. The new performance indicator is the mathematically instrument to measure the continuity of supply and is supported by analytically advantages and disadvantages.The result shows the choice of performance indicators cause a great impact in the revenue cap regulation. AIT, AIF corresponds to ENS and PNS, respectively, divided by power. The three most important results are given by:1. Mathematical and theoretical results show that ENS, PNS are not good indicators individually or in combination with SAIDI, SAIFI and CEMI4 as power consumption within each customer group varies in the local electrical grid.In the regional electrical grid, power consumption may also vary between the norm period and the supervisory period, which means that ENS, PNS can sometimes be misleading.2. The inclusion of power disruption over 12 hours generate stronger incentives in the regulation. Because it requires long term planning to avoid interruptions exceeding 12 hours. The difference is shown mainly in local electrical grid with the indicators SAIDI, SAIFI and CEMI4 and gives a slight increase in rewards in the regulation. This affects the DSO’s and their customers with a marginal difference.3. In the short term, the introduction of AIT, AIF as quality indicators means that customers who consume more power within their customer group get higher-value interruptions. An interruption of a high-consumption customer would then be prioritized compared to a customer with a lower consumption. One way to counteract this is to use CEMI4, in order to capture these customers' interruptions in the regulation.In the long term, the regulation does not become cyclically sensitive, meaning that reasonable rewards or penalties are made. This should benefit a long-term planning of the electrical grid, as the DSO’s do not have to compensate for this. / Det elektriska distributionsnätet är ett naturligt monopol. Infrastrukturen av nätet är uppdelat i områden och där nätägaren har nätkoncession för området som omfattas. Koncessionen är styrd av lagar och reglering.Energimarknadsinspektionen (Ei) är en tillsynsmyndighet som arbetar med uppdrag från regeringen. En del av tillsyn för energimarknaden, är reglering av leveranssäkerhet i det elektriska distributionsnätet. Incitamentet i regleringen ges av intäktsramens begränsning vilket kan resultera i en ökning eller minskning av intäktsramen.Kvalitetsindikatorer är ett matematiskt verktyg för att mäta leveranssäkerhet och används i regleringen av intäktsramen.Examensarbetet syftar till att synliggöra de möjliga konsekvenserna för nätföretagen respektive deras kunder av olika förändringar av de mätmetoder som används för att uppskatta leveranssäkerhet.De mätmetoder som används styrs av indikatorer. Indikatorer som mäter antal avbrott per totalt antal kunder kallas SAIFI. Indikatorer som mäter tid för dessa avbrott per totalt antal kunder kallas SAIDI. En annan metod är att de ingående indikatorerna ska mäta icke levererad energi eller effekt och då kallas de ILE respektive ILEffekt.Energimarknadsinspektionen har inför reglerperioden år 2016-2019 tillämpat SAIDI och SAIFI för kunder kopplade till lokalnät och ILE och ILEffekt för kunder och gränspunkter inom regionnätetExamensarbetet är utformat för att analysera nya typer av kvalitetsindikatorer i lokalnät och regionnät. De nya kvalitetsindikatorerna är olika matematiska verktyg för att mäta leveranssäkerheten och motiveras med analytiska för- och nackdelar.Resultatet visar att valet av indikator har stor betydelse i intäktsramens reglering. AIT, AIF motsvarar ILE respektive ILEffekt dividerat med effekt. De tre viktigaste resultaten ges av:1. Matematiska och teoretiska resultat visar att ILE, ILEffekt är inte bra indikatorer enskilt eller i kombination med SAIDI, SAIFI och CEMI4 då effektförbrukningen inom varje kundgrupp varierar för lokalnät.I regionnät så kan effektförbrukningen också variera mellan normperiod och tillsynsperiod vilket ger att ILE, ILEffekt ibland kan bli missvisande.2. Avbrott över 12 timmar genererar ett starkare incitament i regleringen. Eftersom det kräver långsiktig planering för att undvika avbrott som överstiger 12 timmar. Skillnaden visas främst i lokalnät med indikatorerna SAIDI, SAIFI och CEMI4 och ger en svag höjning av tillägg i regleringen sett på en systemnivå. Det påverkar nätföretagen och deras kunder med en marginell skillnad.3. På kort sikt innebär införandet av AIT, AIF som kvalitetsindikatorer att de kunder som förbrukar mer inom sin kundgrupp får högre värderade avbrott. Ett avbrott hos en kund med hög förbrukning skulle då prioriteras före en kund med lägre förbrukning. Ett sätt att motverka detta är att använda CEMI4 för att fånga upp dessa kunders avbrott i regleringen.På lång sikt innebär det att regleringen inte blir konjunkturkänslig, vilket innebär att rimliga tillägg eller avdrag görs. Det borde gynna en långsiktig planering av elnätet då elnätsföretagen inte behöver kompensera för detta inom regionnäten och lokalnäten.

As our electric power distribution infrastructure has aged, considerable investment has been applied to modernizing the electrical power grid through weatherization and in deployment of real-time monitoring systems. A key question is whether or not these investments are reducing the number and duration of power outages, leading to improved reliability. Statistical methods are applied to analyze electrical disturbance data (from the Department of Energy, DOE) and reliability index data (from state utility public service commission regulators) to detect signs of improvement. The number of installed smart meters provided by several utilities is used to determine whether the number of smart meters correlate with a reduction in outage frequency. Indication emerged that the number of power outages may be decreasing over time. The magnitude of power loss has decreased from 2003 to 2007, and behaves cyclically from 2008 to 2014, with a few outlier points in both groups. The duration also appears to be decreasing between 2003-2014. Large blackout events exceeding 5 GW continue to be rare, and certain power outage events are seasonally dependent. There was a linear relationship between the number of customers and the magnitude of a power outage event. However, no relationship was found between the magnitude of power outages and time to restore power. The frequency of outages maybe decreasing as the number of installed smart meters has increased. Recommendations for inclusion of additional metrics, changes to formatting and semantics of datasets currently provided by federal and state regulators are made to help aid researchers in performing more effective analysis. Confounding variables and lack of information that has made the analysis diffcult is also discussed.

This diploma thesis aims to briefly interpret problems relating to continuity of distribution of electrical energy in a distribution network of low voltage. It also mentions methods of calculation of a distribution system´s reliability which are the most used in practice. Next, there is a chapter dedicated to how distribution networks are projected and operated in the Czech Republic and in the United Kingdom. Based on experiences with meshed networks operation this thesis designs protection of meshed distribution network of low voltage. In the end, a non-sequential method of Monte Carlo is used to calculate System Average Interruption Index (SAIDI) and System Average Interruption Frequency index (SAIFI).

Este trabalho apresenta uma análise crítica da metodologia usada no PRODIST 2010, e propostas alternativas para a fixação de metas e tarifas em função da qualidade do serviço. Todas as análises se sustentam em estudos estatísticos dos dados de interrupções nos anos 2007 e 2008 de 6 empresas distribuidoras de energia com diferentes características de DEC, FEC, número de usuários e número de conjuntos. A primeira proposta chamada de Tarifa Única Meta Única (TAMU) propõe a mesma meta de qualidade do serviço para todos os usuários de um mesmo nível de tensão. Na proposta se fixou como critério geral que 10% dos usuários pior atendidos de um nível de tensão devem ser ressarcidos por má qualidade do serviço. Para alcançar esse objetivo as metas dos níveis de tensão de cada empresa foram fixadas a partir dos seus histogramas de frequência acumulada dos indicadores individuais dos usuários. Esse procedimento garante que a percentagem de usuários ressarcidos seja igual em todos os níveis de tensão das empresas. Os resultados obtidos no cálculo da multa média por usuário ressarcido conservam uma coerência justa com a piora dos indicadores coletivos (isto não acontece no PRODIST), onde as concessionárias que apresentaram os piores desempenhos pagariam uma multa média maior por usuário. A segunda proposta descreve uma metodologia de tarifas diferenciadas por descontos (TARDIF). Nessa metodologia os DIC dos usuários de um mesmo nível de tensão foram classificados em 5 intervalos múltiplos de sua média coletiva (DEC). Os últimos 3 intervalos de classificação correspondem aos usuários que tiveram a pior qualidade do serviço no ano. Para esses usuários foram definidos descontos tarifários que compensariam os prejuízos no seguinte ano. Nos resultados obtidos, as multas médias dos usuários ressarcidos conservam a relação com a média de atendimento e com a dispersão dos dados, sendo o último o critério mais dominante. / This work presents a review of the methodology used in PRODIST 2010, and alternative proposals for setting targets and rates depending on quality of service (QoS). All analysis are supported in statistical interruption data studies of 6 electricity distribution companies (EDC) between 2007 and 2008 with different characteristics of DEC, FEC, number of users and sets. The first proposal called Single Rate - Single Goal (TAMU in Portuguese) proposes the same goal of QoS to all same voltage level users. This proposal sets as a general criterion that 10% of users with worst voltage level attendance must be compensated for getting poor QoS. In order to achieve this goal, the top voltage levels of each company were determined from its cumulative frequency histograms of the user individual indicators value. This procedure guarantees the same percentage of users compensated in all EDC voltage levels. The results obtained by calculating the average penalty per user refunded maintain a fairly consistent with deterioration of collective indicators values (this does not happen in PRODIST 2010), where EDC with worst performances will pay a higher average penalty per user. The second proposal describes a different discount rate methodology (TARDIF in Portuguese). In this methodology, the DIC of users with same voltage level were classified into five intervals of its collective average (DEC). The last three classification intervals correspond to users who have had the worst QoS in the year. For these users were defined discount rates that would compensate the losses next year. In the obtained results, the average penalties of refunded users maintain the relationship between the average attendance and the data scattering, being the last one the most dominant criterion.

Este trabalho apresenta uma análise crítica da metodologia usada no PRODIST 2010, e propostas alternativas para a fixação de metas e tarifas em função da qualidade do serviço. Todas as análises se sustentam em estudos estatísticos dos dados de interrupções nos anos 2007 e 2008 de 6 empresas distribuidoras de energia com diferentes características de DEC, FEC, número de usuários e número de conjuntos. A primeira proposta chamada de Tarifa Única Meta Única (TAMU) propõe a mesma meta de qualidade do serviço para todos os usuários de um mesmo nível de tensão. Na proposta se fixou como critério geral que 10% dos usuários pior atendidos de um nível de tensão devem ser ressarcidos por má qualidade do serviço. Para alcançar esse objetivo as metas dos níveis de tensão de cada empresa foram fixadas a partir dos seus histogramas de frequência acumulada dos indicadores individuais dos usuários. Esse procedimento garante que a percentagem de usuários ressarcidos seja igual em todos os níveis de tensão das empresas. Os resultados obtidos no cálculo da multa média por usuário ressarcido conservam uma coerência justa com a piora dos indicadores coletivos (isto não acontece no PRODIST), onde as concessionárias que apresentaram os piores desempenhos pagariam uma multa média maior por usuário. A segunda proposta descreve uma metodologia de tarifas diferenciadas por descontos (TARDIF). Nessa metodologia os DIC dos usuários de um mesmo nível de tensão foram classificados em 5 intervalos múltiplos de sua média coletiva (DEC). Os últimos 3 intervalos de classificação correspondem aos usuários que tiveram a pior qualidade do serviço no ano. Para esses usuários foram definidos descontos tarifários que compensariam os prejuízos no seguinte ano. Nos resultados obtidos, as multas médias dos usuários ressarcidos conservam a relação com a média de atendimento e com a dispersão dos dados, sendo o último o critério mais dominante. / This work presents a review of the methodology used in PRODIST 2010, and alternative proposals for setting targets and rates depending on quality of service (QoS). All analysis are supported in statistical interruption data studies of 6 electricity distribution companies (EDC) between 2007 and 2008 with different characteristics of DEC, FEC, number of users and sets. The first proposal called Single Rate - Single Goal (TAMU in Portuguese) proposes the same goal of QoS to all same voltage level users. This proposal sets as a general criterion that 10% of users with worst voltage level attendance must be compensated for getting poor QoS. In order to achieve this goal, the top voltage levels of each company were determined from its cumulative frequency histograms of the user individual indicators value. This procedure guarantees the same percentage of users compensated in all EDC voltage levels. The results obtained by calculating the average penalty per user refunded maintain a fairly consistent with deterioration of collective indicators values (this does not happen in PRODIST 2010), where EDC with worst performances will pay a higher average penalty per user. The second proposal describes a different discount rate methodology (TARDIF in Portuguese). In this methodology, the DIC of users with same voltage level were classified into five intervals of its collective average (DEC). The last three classification intervals correspond to users who have had the worst QoS in the year. For these users were defined discount rates that would compensate the losses next year. In the obtained results, the average penalties of refunded users maintain the relationship between the average attendance and the data scattering, being the last one the most dominant criterion.

In today’s society the electric grid has become one of the most important infrastructures and industries as well as other infrastructures, and individuals rely on its functionality. The unavailability of electricity caused by outages is increasingly rare but when it occurs it can have serious consequences. An electric grid’s reliability is now a requirement, not only from customers but also from the regulatory authority.The electric grid in Stockholm is owned and operated by Ellevio AB. Over the past decade the reliability, measured in SAIDI (System Average Interruption Duration Index), in the area has deterioated, mainly due to failures of the medium voltage grid (11- kV, MV). The network structure on the MV- grid consists primarily of two parallel medium voltage cables each of which acts redundantly to its pair cable. The technique, called dual cable structure, works most optimally with substations with automatic switching, which automaticly switch over to the redundant cable when failure occurs. However substations without this automatic switching are widely used resulting in a longer interruption time. Another network structure is the ring or loop structure, which still has redundancy, but when failure occurs a load- break switch must close so that the grid can be fed from another direction. The substation in the loop structure can also be remotely controlled, making the load-break swith operable from the control center, resulting in a shorter interruption time.This thesis evaluates and compares different network structures consisting of dual cable with automatic switching and loop structure with remotely controlled substations, based on economy and reliability, on Stora Essingen. The two structures have been designed in the network information system program Trimble NIS and divided into two different environments; one developed and one undeveloped environment. The results show that the technical solutions in network structures of MV- grid do not necessarily mean major differences in economy nor reliability, rather the choice between few but long interruptions and many but short interruptions is central. Results show however that a combination of the dual cable structure with remotely controlled substations in the undeveloped environment is economically motivated while the reliability can be maintained at the same tame.

The aim of my thesis was to clarify the basic concepts and calculations in the area of reliability of power distribution, as an important element for increase the quality of supplied electric power, this issue we examined in Chapter 3. In the previous chapter, we explained the fundamental solution of MV networks in the Czech Republic and Germany. In Chapter 4 we explained the difficulties in evaluating their values and comparisons between distributional companies, which follows that there should be a uniform procedure for the storage and collection of data for power outages. And the individual distribution companies should follow this standard.