15.07.2015   EuroSun2008-3

Operation of decentralised generation in the distribution grid

The different operation strategies for the cogeneration plants influence the power quality in the grid. To evaluate the power quality we used a load flow analysis [9]. For that analysis it is assumed that all load and generation are 3 phase symmetric. So the result will be the same for all phases. The grid is represented by the admittance matrix, which contains all elements of the used equivalent circuits. The algorithm for solving the load flow equations is implemented in the open source language R (www. r — project. org).

To evaluate the power quality a load flow analysis for the low voltage grid described above was made. For the analysis it was assumed that the electricity demand is distributed equally to the 4 houses. Also the 4 PV plants have the same generation schedule. The feed-in of the CHP was defined by the optimisation results described above. As slack node the node GRID (Fig. 1) was set. At this node the voltage is fixed the reference voltage of 395 V. This is less than the nominal voltage of 400 V to increase the input of decentralised generation in the grid. The slack node also must absorb/provide the needed/surplus power. The load flow analysis was repeated for each time step.

image121

Fig. 6 shows the simulated voltage profile at the regarded day in May at the connection point of the CHP system for five scenarios. In the “BASE” scenario no decentralised generation is available. In this traditional scenario the lowest voltages can be seen. Beginning from the slack node the voltage decreases depending on the height of load until the end of the line. In the scenario “PV only” the four PV plants feed in and change the load in the grid. This results in an increasing voltage during PV input to about 424 V, which almost reaches the upper limit of the allowed voltage range of (+6 %/-10 % of the nominal voltage). In these times additional CHP generation will violate the voltage criterion, which can be seen in the three CHP scenarios. In the scenario “KWK-G” the cogeneration plant is thermal driven and feed in according to the schedule shown in Fig. 3. It can be seen clearly that some of the operation blocks during high PV input times increase the voltage up to 435 V which violates the voltage criterion. A similar behaviour can be seen in the scenario “VDE” The highest payment for

produced energy is at 10 am where the transport capacity is restricted due to high PV input. In “LOCAL” scenario the high price times are in the early morning and evening. So the CHP operation blocks are shifted to these times. Because of this shifting all the produced energy of the CHP can be used in the local grid and no further violations are caused. The maximum voltage does not exceed the maximal voltage in the “PV only” scenario.

As it can be seen in the results of Fig. 6 the locally optimised CHP operation with a local price curve can help to control decentralised generation with respect to restrictions from the grid. In the scheduling process of generation it must be avoided, that all generators feed in simultaneously if there is little load and the grid is near its transport capacity. Because almost all plants do have local conditions to be fulfilled (e. g. heat restriction for CHP plants) it seems not to be possible to control all of them by a central device. These restrictions only can be known by the local operator. With the local price curve he also will automatically act according to the grid operator’s interests as much as he can.

2. Conclusion

Within this paper we demonstrate the potential of local management of decentralised generation to increase grid capacity towards RES and DER, taking cogeneration as example. For times of high grid utilization due to local excess energy production, lower pay-back prices for controllable electricity feed-in are defined by the local grid operator, while during times of low grid load prices are increased accordingly. This local component to the price curve for produced electricity stimulates local operators to shift their generation to high-price times. With this control method the grid operator can influence operation without direct access to cogeneration devices and the amount of RES can be increased without upsizing the grid facilities. This method can easily be extended to loads, which will be influenced by flexible tariffs. With the propagation of “Smart Metering” systems, the option of tariff driven loads and generation will be available in the near future.

References

[1] A European Strategic Energy Technology Plan: Technology Map, Commission of the European Communities, SEC(2007)1510, Brussels 2007

[2] Leprich, U.; Bauknecht, D.: Dezentrale Energiesysteme und aktive Netzbetreiber, BMU-Fachtagung Perspektiven Dezentraler Energiesysteme, Berlin 2006

[3] www. netmod. org

[4] Wille-Haussmann, B., Erge, T., Wittwer, C.: Decentralised Optimisation of Cogeneration in Virtual Power Plants, CISBAT 2007 — Renewables in a changing climate Innovation in the built environment, Lausanne, Switzerland, 4.-5.9.2007

[5] Bronstein, I.: Taschenbuch der Mathematik, Harri Deutsch 1999

[6] VDE-Studie: Smart Distribution 2020: Virtuelle Kraftwerke in Verteilungsnetzen, 2008

[7] Backhaus, K., Erichson, B., Plinke, W., Weiber, R.: Multivariate Analysemethoden, Springer 2005

[8] Faraway, J.: Practical Regression and Anova using R, 2002

[9] Spring, E.: Elektrische Energieverteilnetze, VDE-Verlag 2003