MODEL FOR LOSSES CALCULATION AND BREAKDOWN IN DISTRIBUTION SYSTEMS. CONCLUSIONS

Transcription

1 MODEL FOR LOSSES CALCULATION AND BREAKDOWN IN DISTRIBUTION SYSTEMS. R O FERREYRA P J PAOLETICH EDEA SA - ARGENTINA SUMMARY We are presenting a model for losses calculation in distribution system, trying to make it simple, relatively easy to implement and sufficiently accurate. The model is based on: The incoming energy through the boundary of the Distribution Company, with the Transmission Companies and Generators. Versus The energy sold to its own customers plus power transmitted for agents of the Wholesale Electrical Market (WEM) in its concession area. The difference between them represents the total energy lost. In order to calculate the technical losses and to be able to break them down from the total, the model considers the voltage level of each of the nodes through which the power enters the concession area and the voltage level of the point where the power is sold or received by a Market Agent. This is so, in order to follow the power flow through the different stages of the system, allocating the corresponding technical losses to each of them. The stages of the distribution system that have been considered are the following: á Transmission through the 132 Kv network á Transformation 132/33 Kv á Transformation 132/13.2 Kv á Transmission through the 33 Kv network á Transformation 33/13.2 Kv á Distribution in 13,2 Kv á Transformation 13,2/0.4 Kv á Distribution in Low Voltage A percentage of technical losses is allocated at each stage, as accurately as possible, according to the available data and the calculation tool. As we go down in the distribution system, the study of technical losses becomes more complex. Therefore, the 132 Kv network, is easier to analyse than the low voltage network, as load information is generally available at all nodes and year round. Including hour by hour data, which accurately provides essential parameters, such as the utilisation time and the equivalent time in losses. In addition, this network is analysed through the use of tools for the load flows calculation at different stages, enabling the calculation of the percentage of losses with acceptable accuracy. On the other hand, the extent of low voltage network prevents its complete study enabling only a statistic analysis. Finally, we are presenting a sensitivity analysis to evaluate which stage of the distribution system is more significant to determine the technical losses in order to focus the efforts on the best value. Paper Recipients This paper is intended for the staff working for Distribution Companies with high indexes of total losses, who must analyse technical losses in order to differentiate them from the total. The initial information requirement is quite basic, being able to start by using usual values from similar companies and gradually adjust their own indexes. CONCLUSIONS The methodology described has been applied in EDEA S.A. since July 1997 with satisfactory results. However, sub-system indexes must be frequently revalidated. Accuracy is expected to improve in the future by the existence of more powerful calculation tools and greater load information availability.

2 MODEL FOR LOSSES CALCULATION AND BREAKDOWN IN DISTRIBUTION SYSTEMS. R O FERREYRA P J PAOLETICH EDEA SA - ARGENTINA SUMMARY We are presenting a model for losses calculation in distribution system, trying to make it simple, relatively easy to implement and sufficiently accurate. The model is based on: The incoming energy through the boundary of the Distribution Company, with the Transmission Companies and Generators. Versus The energy sold to its own customers plus power transmitted for agents of the Wholesale Electrical Market (WEM) in its concession area. The difference between them represents the total energy lost. In order to calculate the technical losses and to be able to break them down from the total, the model considers the voltage level of each of the nodes through which the power enters the concession area and the voltage level of the point where the power is sold or received by a Market Agent. This is so, in order to follow the power flow through the different stages of the system, allocating the corresponding technical losses to each of them. The stages of the distribution system that have been considered are the following: á Transmission through the 132 Kv network á Transformation 132/33 Kv á Transformation 132/13.2 Kv á Transmission through the 33 Kv network á Transformation 33/13.2 Kv á Distribution in 13,2 Kv á Transformation 13,2/0.4 Kv á Distribution in Low Voltage A percentage of technical losses is allocated at each stage, as accurately as possible, according to the available data and the calculation tool. As we go down in the distribution system, the study of technical losses becomes more complex. Therefore, the 132 Kv network, is easier to analyse than the low voltage network, as load information is generally available at all nodes and year round. Including hour by hour data, which accurately provides essential parameters, such as the utilisation time and the equivalent time in losses. In addition, this network is analysed through the use of tools for the load flows calculation at different stages, enabling the calculation of the percentage of losses with acceptable accuracy. On the other hand, the extent of low voltage network prevents its complete study enabling only a statistic analysis. Finally, we are presenting a sensitivity analysis to evaluate which stage of the distribution system is more significant to determine the technical losses in order to focus the efforts on the best value. Paper Recipients This paper is intended for the staff working for Distribution Companies with high indexes of total losses, who must analyse technical losses in order to differentiate them from the total. The initial information requirement is quite basic, being able to start by using usual values from similar companies and gradually adjust their own indexes. THE MODEL EDEA S.A. has boundary points with the Transmission Company TRANSBA in 132, 33 and 13,2 Kv in 20 Transformer Stations and with ESEBA GENERATION in the 9 de Julio of Mar del Plata Power Plant in 132 Kv. The total boundary metering includes 78 metering points of the Commercial Metering System, SMEC. Part of the energy that goes through the boundary is sold or supplied to GUMAS (Major Large customers, WEM agents) at the same purchase node. Therefore, the transfer occurs without significant technical losses. The remaining energy will wear away as it goes into the system, either as technical losses, sales or transmission for market agents. Finally, the remaining energy that cannot be allocated to any of the aforementioned categories constitutes the non-technical losses. Figure 1 shows the model proposed for EDEA S.A., and the letters A, A3 and H2 indicate the power that crosses the boundary in a certain time unit, in 132, 33 and 13,2 Kv voltages respectively.

3 If we follow the energy flow from the first 33 Kv bus bar, it is fed by the sum of all the boundary readings at that voltage level, plus that transformed to that voltage after crossing the boundary at a different voltage. If we deduct at this point the energy sold at the (B13) boundary point and the one delivered at this point to GUMAS (C4), only the difference (D3) will produce losses in the 33 Kv network according to the percentage allocated (E3). In equations: Energy flowing through the 33 Kv network. D3 = A3+U-(B13+C4) Technical losses calculated in the 33 Kv network E3 = l 33 %. D3 Continuing downstream, it will be transformed to 13,2 Kv with the consequent transformation losses (G3), only a fraction of the power that entered the 33 Kv network, to which we must deduct the losses (E3), the sales in the 33Kv network (B23) and the transmission through the 33Kv network (C3). In equations: Energy transformed from 33 to 13,2 Kv F3 = D3-(E3+C3+B23) Technical losses calculated in 33/13.2 Kv transformation G3 = t 3/13.2 %. F3. The remaining energy (H3), enters the 13,2 Kv system, where it is added to the one that enters through the boundary points at that voltage (H2), and to the one that having entered in 132 Kv network was not deducted on the way as sales, transmission or losses (H1). Following this reasoning until reaching the low voltage, after deducting the sales and losses at that voltage, the surplus energy, as we already said, constitutes the non-technical losses. These simple equations of additions, subtractions and multiplication are presented on a spreadsheet, like the one shown in figure 2 in accordance with the scheme of figure 1. There, we can see the energy flowing through each section of the network and the respective losses in absolute values and in percentage values referred to the total energy entering the system. As we already said, part of the energy that enters is delivered at the same node without causing technical losses. In consequence the percentages referred to the energy total are not indicative of the network efficiency, therefore, in the lower part of the diagram, losses are referred to the power that actually enters the network. 132Kv 33Kv TRANSBA C4 C3 A3 D3 F3 B13 E3 B23 G3 H3 GENERACIÓN ~ T U V C H3 H2 J1 13,2Kv J2 P3 0,4Kv A D E B1 F B2 G H1 K I1 L L1 I2 M N O P1 Q R P2 S Figure 1 Strengths and Weaknesses of this model Model application is relatively simple. Any company has available data on purchased, transmitted and sold energy; it only needs to incorporate the voltage level with which the transaction is made. The model takes account of the power transmitted for other market agents, since they also cause technical losses.

4 In principle, non-technical losses will occur at the low voltage level, but part of them will possibly originate at other voltage levels, automatically reducing the ones at low voltage and the portion of technical losses that these provoke. The main weakness and challenge of the model is to have correct technical losses percentages for each stage in the system. These percentages must be adjusted every time there are significant changes in the network s configuration or in the demand parameters. ENERGY FLOW - EDEA S SYSTEM - MARCH 2000 GWh Incoming Energy 132 system A 103,01 53% Total Energy sales in 132 B 10,99 6% Energy billed in Transba s busbars B1 10,99 6% Other sales in 132 B2=B-B1 0 0% Gumas in 132 C 0% Energy flow in 132 D=A-B1-T 85,79 44% Losses calculated in 132 E 0,86 0% Energy Transformed to 13.2 F=D-E-C-B2 84,93 44% Transformation Losses 132/13.2 G 0,42 0% Energy passing to 13.2 system H1 84,51 44% Energy purchased in 33 A3 51,62 27% Contrib. 33 System from MDP U 6,20 3% Incoming E. 33 system A3+U 57,82 30% Losses transformation 33 MDP V 0,03 0% 132 transformed to 33 in MDP T= U+V 6,23 3% Total Energy sales in 33 B3=B13+B23 17,84 9% Energy billed Transba s busbars B13 16,33 8% Other sales in 33 B23 1,51 1% Gumas in 33 C3 1,00 1% Gumas in 33 on Transba s busbars C4 1,46 1% Energy flow in 33 D3=A3+U-B13-40,03 21% Losses calculated in 33 E3 1,15 1% EnergyTransformed to 13.2 F3 36,37 19% Losses Transformation 33/13.2 G3 0,36 0% Energy passing to 13.2 system H3 36,01 19% Energy Purchased in 13.2 H2 39,62 20% Incoming energy 13.2 system H=H1+H2+H3 160,14 82% E. Total sales 13,2 Kv I 52,88 27% Energy billed in 13.2 bars I1 37,95 20% Other sales in 13.2 I2 14,93 8% Gumas in 13.2 bars J1 7,78 4% Gumas in 13.2 network J2 3,55 2% Energy flow in 13.2 K 114,41 59% Losses calculated in 13.2 Kv L 1,14 1% Non-technical losses in 13.2 Kv L1 0,5 0% Power Transformed to L.V. M 94,29 49% Losses Transformation 13.2/0,4 N 1,98 1% Incoming energy L.V. network O 92,31 48% Total energy sales L.V. P 82,52 42% Energy billed L.V.busbars P1 10,00 5% Other sales in L.V. P2=P-P1 72,52 37% Gumes in LV P3 1,44 1% Energy flow in L.V. Q 80,87 42% Losses calculated in L.V. R 1,21 1% Non-technical losses in LV S 7,13 4% Total Technical Losses 7,16 4% Total Losses 14,80 8% Total Gumas/es 15,23 8% NETWORK EFFICIENCY Power through EDEA s System 119,75 100% Technical Losses 7,16 6% Non-technical losses 7,63 6% Total Losses 14,80 12% PERCENTAGE OF TECHNICAL LOSSES ALLOCATED IN EDEA SA TO EACH STAGE OF THE DISTRIBUTION SYSTEM. Transmission through the 132Kv network:1.0 % EDEA S.A. s 132 Kv network is not very large and mostly underground. For this reason, the percentage allocated is low as compared to the other companies with large overhead lines. The value was obtained from the calculation of load flows for the different peak, valley and rest times during summer and winter, subsequently affected by the corresponding times and losses factor. Transformation 132/13.2 Kv: 0.5 % The percentage was determined through the analysis of the performance curves of the sets in service, which are practically flat within the hourly and seasonal variation rating of the load. Figure 2 shows the curve of a 44 MVA transformer. Losses [%] 10,0 9,0 8,0 7,0 6,0 5,0 4,0 3,0 2,0 1,0 Losse Percentage in Transformers 0,0 90,0 0,00 0,20 0,40 0,60 0,80 1,00 1,20 S/Sn Losses [%] Performance [%] Figure 2. Transformation 132/33 Kv: 0.5 % We adopted the value determined for the 132/13.2 Kv transformers. The little power affected by this transformation is in 123/33/13.2 Kv three-winding transformers. 100,0 Transmission through 33 Kv network: 2.87 % Using existing data from capacity records every 15 of interurban lines that support the 50 % of the total power transmitted at that voltage level, we obtained that the losses represent 2.87 %, being 0.48 and 6.76 % the extreme values. 99,0 98,0 97,0 96,0 95,0 94,0 93,0 92,0 91,0 Performance [%]

5 Transformation 33/13.2 Kv: 1 % Calculated for the 132 Kv transformers, with the particularity that the sets at this voltage level have inferior performance, as they have lower capacity. Where: Ep n : Energy lost in n transformer En : Energy demanded in n transformer Distribution in 13,2 Kv 1.0 % This percentage was taken form the international standards of countries in the region [1], which locate them between 0.90 and 1.03 %. In the future, we expect to obtain more accurate value by using, data from company ongoing projects, which will enable to export the network topology to capacity flow calculation tools, complementing it with data from telecontrol. Transformation 13,2/0.4 Kv 2.1 % From the production of 15-day records at 60 Substations, taken at random. Adding the totality of losses, fixed and variable, and relating them with the total energy delivered during the same period, we obtained the indicated percentage. Even though the under-loaded sets, generally, of small capacity, present a high percentage of losses, these are not significant in absolute value, reason why they are not preponderant in the final result. Graph 3 shows accumulated losses and demand values, in such a way that the relation between losses and demand of the last point represents, per unit, the losses of the whole group. Distribution in Low Voltage: 1.50% Calculated with the support of Network planning Software, based on typical Substations demand records and knowledge on the network of the area. For each line, we considered the utilisation time, the losses equivalent time, and a factor that represents the losses increase due to unstable load between the phases. The maximum demand value was distributed in the network proportionally to the length of each section of the line, separated with a distance of approximately 10 mts. Finally, with the losses capacity obtained as a result of the capacity flow run, and the factors already mentioned, we calculated the lost and the demanded power in the area supplied by each Substation. Accumulated losses [MWh] LOSSES IN DISTRIBUTION TRANSFORMERS ,000 Accumulated Demanded Energy [Mwh] Figure 4 The number of case studied so far is not sufficient to draw conclusions, but it is very close to the value historically adopted. It is important to make clear that 0.5 % will be added to the value resulting from the study, which is the value coming from the losses shown in the meter voltage coils, approximately of 0.7 W, connected 8760 hours per year, as regards the power volume operated in low voltage, which is approximately 48% of the total. SENSITIVITY p[ / 1] = Figure 3. Ep1 + Ep Ep E1 + E En n Even though subsystem losses percentage may vary from one to another, their quantity of energy flow determines each sub-system weight for the calculation of total technical losses. Starting from an initial situation in which the percentage of losses at each stage is the one considered as most

6 acceptable, which we will call stage 1. We then increase the percentage by one point in one of them and obtain stage 2 for the sub-system under analysis, where the percentages of the rest of the network remained constant. Having done this for each of the network defined elements, and showing them in a graph as the one in Figure 5, we see that the elements represented in the top curve have greater influence on the final result than the ones in the lower curve. SENSITIVITY ANALYSIS Effect of increasing, by one point, the losses of an element over the total percentage of losses CONCLUSIONS The methodology described has been applied in EDEA S.A. since July 1997 with satisfactory results. However, sub-system indexes must be frequently revalidated. Accuracy is expected to improve in the future by the existence of more powerful calculation tools and greater load information availability. Technical losses (%) 7,00% 6,80% 6,60% 6,40% 6,20% 6,00% 5,80% 6,92% 6,78% 6,72% 6,68% 6,32% 6,30% Stage 1 Stage 2 L ,00% 6,92% Tr. 13.2/0.4 6,00% 6,78% Tr. 132/13.2 6,00% 6,72% L ,00% 6,68% L. B.T. 6,00% 6,68% L Tr. 13.2/0.4 Tr. 132/13.2 L. 132 L. V. L. L. 33 Tr. 33/13,2 Stage 2 considers a one point increment in the losses percentage as regards stage 1 Figure 5 Particularly, in our company, the 132 Kv network is the most sensitive one. On the other end, the least sensitive one are the 33Kv lines and transformers. The aforementioned should not be taken as a general conclusion, but as one that obeys to the particular configuration of EDEA s market, schematically shown in figure 6. [1] Mario Luis Martín, Orlando Héctor Ramati. Pérdidas de Energía En la Distribución. Comisión de Integración Eléctrica Regional CIER Uruguay [2] M.V. Gonzalez Sábato, Saverio Scioscia. Evaluación y Medición de pérdidas eléctricas técnicas y no técnicas. Políticas y actividades para la disminución de las pérdidas no técnicas. Experiencia en ESEBA S.A. CIRED Buenos Aires KV 33/ KV 13.2/ KV 132/13.2 Scheme of the power volumes operated in each of the network sub-systems. Figure 6