oa Dynamic Tariff Development for Effective Demand Side Management (DSM) in the Presence of Increased Penetration of Photovoltaics (PV)

I. Introduction

The trend towards high penetration of renewable energy sources (RES) in the energy mix and particularly grid-connected photovoltaic (PV) systems in the low voltage (LV) network, offers the benefits of green decentralized generation, at the cost of the development of energy management tools to alleviate potential problems. More specifically, the fact that for most consumption profiles the PV energy production does not coincide with the electricity demand, forces the grid to act as a sink and a source thus requiring re-adaptation of the grid operation [1]. To this extent, an advanced demand side management (DSM) scheme can be introduced to mitigate RES operational issues and contribute to managing effectively congestion problems. In this work, a price-based DSM tool has been developed in order to arrive at an effective Time of Use (ToU) tariff with improved DSM results. In this scope, smart meters (SMs) have been deployed at three hundred households with grid-connected PV systems installed at their rooftops, in order to acquire consumption and production profile details of typical Cypriot prosumers. The selected pilot sites that already have a 3 kWp grid-connected PV system, have been offered a ToU tariff allowing rates and charges to vary based on the time and date of consumption, i.e. day or night and seasonality. This aims to examine how financial implications can alter the energy behavior of prosumers [2]. In addition, the prosumers are divided into three groups each receiving a distinct method of monitoring their energy habits: one group will have In-House Displays (IHDs), the other will have access through a web application, and the third will receive information about their energy habits through the traditional bi-monthly mail bill. This will facilitate an in-depth examination of the prosumers’ response based on the information sent to them [2].

II. Background

Theory Price-based DSM programs offer an alternative to the traditional flat tariffs and comprise of Critical Peak Pricing (CPP), Real-Time Pricing (RTP) and Time-of-Use (ToU) tariffs [3], [4]. Amongst the different schemes, ToU tariffs are commonly preferred because the price of energy consumption is fixed for different periods of the day in contrast to other price-based DSM programs where the price fluctuates following the real time cost of electricity [3], [4]. The development of an efficient DSM system offers the advantage of generating cost reductions for grid utilities and the increase of operational efficiency. However, in order to achieve a balanced DSM scheme, the existing flat tariffs need to be transformed to ToU tariffs providing the necessary monetary incentives for domestic consumers to flatten their load profile. Even though, ToU tariff schemes offer the advantage of price certainty, the effectiveness of such tariff schemes must be verified prior to implementation because of the risk of a new peak appearing through load shifts at cheaper price periods, posing negative effects on the optimal operation of the system [5]–[8].

III. Prosumer Features

In support of this work, three hundred prosumers in Cyprus have been selected through the implementation of the SmartPV project (http://www.smartpvproject.eu/), in order to acquire real-time information of the consumption and production profiles and to identify the potential problems and limitations of the existing energy policy. All participating prosumers are geographically spread in Cyprus, in order to cover different socio-geographical conditions and thus targets a variety of consumers. In addition, prosumers with higher total yearly electricity consumption (in kWh) compared to the typical energy production from a 3 kWp grid-connected PV system were selected.

IV. Methodology

The development of a dynamic ToU tariff tool to enable price-based DSM relies strongly on the analysis of the basic input parameters such as electricity demand and PV electricity production profiles. Consumption and production data acquired from the three hundred prosumers was used to optimize the dynamic ToU tariff algorithm for the case of Cyprus. It has to be pointed out that this is a benchmark tool which can be used by any other country. The first step in the development of the algorithm was to identify the maxima and minima power consumption periods of the provided average domestic consumption profiles, an approach already implemented and verified [9]. The load duration curve of the provided average domestic consumption profile in Cyprus for each season was analyzed in order to identify possible inflection points. The different inflection points of the curve represent the various load segments which were used in order to obtain the probability density function (PDF) (at a 95% confidence interval) [1]. The PDF of each segment represented the ToU block period. The dynamic tariff tool developed from the above statistical analysis is capable of deriving the ToU blocks with a mean absolute percentage error (MAPE) and root mean square error (RMSE) between the ToU block periods and the load profile of 8.65% and 19.95%, respectively [1]. In order to further improve the initial algorithm, a function based model was developed. Optimization methods were used in combination with the statistical results [1]. In this approach the statistical output ToU block periods are used as the initial condition of the optimization procedure using the Matlab Optimization ToolboxTM. The ToU blocks are directly compared with the load profile rather than extracting the ToU blocks from the load duration curve. The objective function of the optimization procedure minimizes the RMSE as described by (1): where is the derived ToU block period, is the load profile and is the total sampling interval. Based on this equation, is the variable to be optimized and changes according to the desired levels. To achieve this, the developed optimization tool uses a hybrid optimization function such as simulated annealing [10] and pattern search [11]. In summary, two different methods were used to derive the final ToU block periods: a) combining statistical analysis using the load duration profile and b) optimization methods applied to the load profile.

V. Results

A. Dynamic ToU Tariff tool A software application tool was developed in order to assist users to visualize the impact on their electricity bill from the different ToU blocks. The derived ToU blocks for the winter period were compared using two different approaches (figure 1). The MAPE and RMSE between the load curve and the ToU blocks were improved by utilizing the optimization tool reducing them by 2.43% and 7.63%, respectively when compared to the statistical approach. Furthermore, the optimization approach clearly demonstrated that the peak consumption period is charged with the higher tariff, while the lowest tariff occurs during the valley period. Through this approach another period is clearly identified representing the transitional period: from the minima to the maxima and vice-versa. These time periods are important as they can be used by prosumers to cover their needs that can be shifted from the peak periods but cannot wait until the off-peak period (e.g. cooking, devices without smart control etc). B. Acquired data from three hundred prosumers Data-sets collected during the summer period from the three hundred prosumers equipped with Smart Meters (SMs), before the application of ToU tariffs, were analysed in order to evaluate the self-consumption index. Figure 2 presents the consumption and PV production profiles from the pilot sites. The evaluation of the self-consumption energy was calculated using the equations below [12] and the results are presented in Table I. The average self-consumption energy for the participating prosumers during the summer months is calculated to be 738.87 kWh hence, 53.52% of the energy produced is directly consumed on site while the remaining energy is exported to the grid. Comparing this with the typical prosumer of Cyprus, the self-consumption rate is more than 10% higher. This is mainly due to the fact that the SmartPV sample has on average a higher load demand when compared to the typical prosumer.

VI. Conclusions

In this work, the application of DSM schemes aiming at raising the awareness of the consumers, coupled with financial incentives has been demonstrated through dataset collection from 300 prosumers. In this domain, a new tool for evaluating the dynamic ToU tariffs has been developed based on two different methods, in order to promote effective price-based DSM practices in the electricity network of Cyprus. This is based on statistical analysis of the provided average consumption profiles and optimization procedures, aiming to derive the most appropriate ToU tariffs. The statistical method showed a MAPE and RMSE of 8.22% and 19.95%, respectively, by comparing the resulted ToU blocks to the load profile. On the other hand, the optimization method resulted in a MAPE and RMSE of 6.22% and 12.32% respectively, proving its effectiveness and improved accuracy. In addition, energy data-sets have been collected from participating prosumers before the implementation of the ToU tariff and comparisons between measured data and a typical prosumer was made. The results indicated an average self-consumption index with the existing net metering scheme for the summer months of about 53%.

Acknowledgment

The project is co-financed by the program LIFE (LIFE+ Environment Policy and Governance) of the European Union under the grant agreement number LIFE 12/ENV/CY/000276.

References

[1] N. Philippou, M. Hadjipanayi, G. Makrides, V. Efthymiou, and G. E. Georghiou, “Effective dynamic tariffs for price-based Demand Side Management with grid-connected PV systems,” in PowerTech, 2015 IEEE Eindhoven, 2015, p. 5.

[2] B. Davito, H. Tai, and R. Uhlaner, “The smart grid and the promise of demand-side management,” 2010.

[3] M. H. Albadi and E. F. El-Saadany, “A summary of demand response in electricity markets,” Electr. Power Syst. Res., vol. 78, no. 11, pp. 1989–1996, Nov. 2008.

[4] B. Severin, M. Jaske, and A. Rosenfeld, “Dynamic Pricing, Advanced Metering, and Demand Response in Electricity Markets.” Center for the Study of Energy Markets, p. 103, 2002.

[5] S. Gottwalt, W. Ketter, C. Block, J. Collins, and C. Weinhardt, “Demand side management—A simulation of household behavior under variable prices,” Energy Policy, vol. 39, no. 12, pp. 8163–8174, Dec. 2011.

[6] J. Torriti, “Price-based demand side management: Assessing the impacts of time-of-use tariffs on residential electricity demand and peak shifting in Northern Italy,” Energy, vol. 44, no. 1, pp. 576–583, Aug. 2012.

[7] J. Thumim, “Investigating the potential impacts of Time of Use (TOU) tariffs on domestic electricity customers Report to Ofgem,” 2014.

[8] C. M. Flath, “An optimization approach for the design of time-of-use rates,” in IECON - 39th Annual Conference of the IEEE Industrial Electronics Society, 2013, pp. 4727–4732.

[9] Q. Hui-ting and L. Yi-jie, “Research of Peak and Valley Period Partition Approach on Statistics,” in 4th International Conference on Electric Utility Deregulation and Restructuring and Power Technologies (DRPT), 2011, pp. 1774–1779.

[10] L. Ingber, “Adaptive simulated annealing (ASA): Lessons learned,” Invit. Pap. to a Spec. issue Polish J. Control Cybern. “Simulated Annealing Appl. to Comb. Optim., 1995.

[11] T. G. Kolda, R. M. Lewis, and V. Torczon, “A generating set direct search augmented Lagrangian algorithm for optimization with a combination of general and linear constraints,” 2006.

[12] O. Schaefer, S. Willborn, S. Goeke, J. A. Toledo, V. Cassagne, and A. Roesch, “Self Consumption of PV Electricity,” EPIA, 2013.