Case Study. Utility-Owned Smart Grid Smart Grid, Smart Cities Trial. 1 March 31 May 2012

Transcription

1 Case Study Utility-Owned Smart Grid Smart Grid, Smart Cities Trial 1 March 31 May July 2012

2 Background The Australian Government has chosen Ausgrid to lead a $100 million initiative across five sites in Newcastle, Sydney and the Upper Hunter regions (see Figure 1). Smart Grid, Smart City creates a testing ground for new energy supply technologies. At least 30,000 households will participate in the project over three years. The demonstration project gathers information about the benefits and costs of different smart grid technologies in an Australian setting. Building a smart grid involves transforming the traditional electricity network by adding a chain of new, smart technology. It includes smart sensors, new backend IT systems, smart meters and a communications network. RedFlow has won the bid to supply 61 energy storage systems (ESS) to Ausgrid for installation in Newcastle, Scone and Newington in Sydney. 40 of these were installed in Newcastle in late 2011 and early 2012, with the other 20 ESS installed in Scone in April One ESS has also been operational in Newington since 2010 as part of the ongoing Smart Home sub-project. The 40 systems in Newcastle have been operating since February 2012 and have been feeding into the grid during peak demand periods. The 20 systems in Scone have been operational since mid-may RedFlow ESS 40 RedFlow ESS 1 RedFlow ESS Figure 1: The locations and highlights of the Smart Grid, Smart City Trial [1] RedFlow s ESS are based on its core zinc bromide module (ZBM) flow battery technology. Each ESS contains one ZBM, battery management system (BMS), remote terminal unit (RTU), inverter and 3G modem for communications. All components are housed in a metal enclosure installed near customer houses on private property. 11/10/2012 Page ii

3 Summary Increases in both general and peak electricity demand, the integration of intermittent and distributed generation, and developments in communications technology have all enabled, as well as necessitated a more advanced electricity network. This moves away from the traditional unidirectional grid, with central generation and one-way communications. As such, utilities have seen Smart Grids as an answer to the need for greater grid capacity, as well as allowing transmission and distribution infrastructure upgrade deferral. According to the Electric Power Research Institute (EPRI), a Smart Grid integrates and enhances other necessary elements including traditional upgrades and new grid technologies with renewable generation, storage, increased consumer participation, sensors, communications and computational ability [2]. The use of energy storage in Smart Grids can provide utilities with many benefits, including improved operational efficiency and increased value of distributed generation, thereby improving customer satisfaction [3]. This is done through the time-shifting of electricity from low demand, as well as times of high distributed generation output, to high demand times. However, despite the many benefits of energy storage, existing proven technologies are either inappropriate for Smart Grid applications (in the case of lead acid batteries), or are physically impossible to implement on a large scale in a variety of areas (in the case of pumped hydro or compressed air energy storage (CAES)). As such, new and advanced technologies are emerging to satisfy this new market. These must be trialed and tested to ensure that they are both reliable and align with the needs of utilities. As such, the Australian Government has initiated, and is currently funding the Smart Grid, Smart Cities (SGSC) Trial. It aims to test several hypotheses about many types of Smart Grid technology. In particular, the Trial aims to quantify the following seven benefits of energy storage: 1. Reduction in peak demand energy storage as a cost-effective and reliable alternative to network capacity expansion 2. Improvement in network reliability/voltage/power factor/power quality costeffectiveness and value of energy storage 3. Energy supply during peak price events net benefit to retail sector 4. Minimisation of customers energy bills with innovative tariffs e.g. time of use together with energy storage. 5. Combined benefit between consumer, retail and network sectors 6. Investigation of large capacity (~1MVA) storage extra cost and other benefits compared to smaller capacity storage 7. Intermittent generation support optimisation of renewable energy sources value RedFlow has been selected as one of several advanced energy storage companies to supply energy storage systems (ESS) for the trial to test these seven benefits of energy storage to Smart Grids. The ESS supplied for the Trial by RedFlow are the zinc-bromide module (ZBM)-based R510 model (see Appendix B R510 Product Brochure). These systems are rated at 5kW, 10kWh, and comprise one ZBM, inverter, 3G modem for communications, battery management system (BMS), remote terminal unit (RTU) and other power electronics housed in a metal enclosure (see Figure 8). 11/10/2012 Page iii

4 kwh % Power Electronics Including BMS, RTU, 3G Modem Inverter SMA Sunny Backup Enclosure 1850 H x 971 L x 567 W ZBM Figure 2: Components of RedFlow's R510 ESS 61 R510 ESS have been installed and commissioned on separate residential properties as part of the SGSC Trial. The ESS have been staggered in their installations, with all systems operational since mid-may The operational results and benefits of RedFlow s ZBM-based ESS were analysed, and the lessons learnt presented. This produced the following key conclusions: Over the period between 1 March and 31 May 2012, the R510 ESS installed for the SGSC trial have outputted a total of MWh to the grid at an average efficiency of 58.03%. This is shown in the graph below in Figure SGSC ESS Grid Interaction March April May Grid Import (kwh) Grid Export (kwh) Efficiency (%) Figure 3: Total grid import and export over the trial period with average monthly efficiency 11/10/2012 Page iv

5 Power (kva) Grid Power (kw) The use of RedFlow s ESS reduces the traditional peak seen by the grid by 10-15% when used in a ratio of approximately 1 ESS for 16 customers during cooler months (see Figure 4) :40 4:10 4:40 Grid-Feed Period Monday 9 April 2012 Peak Reduction 15.19% 5:10 5:40 6:10 6:40 7:10 7:40 8:10 65 Residential Customers 4 R510 ESS 8:40 9:10 9:40 10:10 Real Power (kw) with Storage Real Power (kw) without Storage Figure 4: Results from the Melinda Avenue feeder in Newcastle show a reduction in peak demand of over 15% with the use of RedFlow's ESS during traditional peal times The use of RedFlow s ESS transforms the traditional evening peak seen by the grid to a trough when used in a ratio of approximately 1 ESS for 1 customer during cooler months (see Figure 5) Power on Miranee Road Recloser Energy Storage in Use Grid Charge Period 28kW Grid Charge Period 12kW Grid- Feed Period 45kW 13 Residential Customers 15 R510 ESS 12:00 AM 3:00 AM 6:00 AM 9:00 AM 12:00 3:00 6:00 9:00 12:00 AM Figure 5: Results from the Scone feeder in mid-june show a transformation from peak to trough with the use of RedFlow's ESS during traditional peak times Research and Development already undertaken by RedFlow shows that its kw-scale ESS can be scaled up to MW-scale systems (see Figure 6). 11/10/2012 Page v

6 Figure 6: RedFlow's M90 90kW, 180kWh ESS was installed at the University of Queensland in late May 2012 RedFlow s ESS perform well when integrated with intermittent solar generation, as well as more reliable fuel cell distributed generation 2 Solar 12 Solar 3 Solar 3 Solar 15 Solar 5 Solar Figure 7: The SGSC trial in Newcastle has integrated a total of 200kW of solar and 50kW of Blue Gen fuel cell generation with 40 RedFlow ESS 11/10/2012 Page vi

7 Contents Background... ii Summary... iii Contents... vii List of Figures... viii 1 Introduction Operation Newcastle Scone Newington Results Newcastle Peak Demand Reduction Reliability Scone Peak Demand Reduction Reliability Newington Load Following Reliability Lessons Learnt Newcastle Scone Newington Conclusions References Appendix A List of Abbreviations Appendix B R510 Product Brochure Appendix C List of Trial Sites Appendix C1 List of Newcastle Sites Appendix C2 List of Scone Sites Appendix D Cycle Profiles Appendix E Newcastle Reduction in Peak Demand /10/2012 Page vii

8 List of Figures FIGURE 1: THE LOCATIONS AND HIGHLIGHTS OF THE SMART GRID, SMART CITY TRIAL [1] II FIGURE 2: COMPONENTS OF REDFLOW'S R510 ESS IV FIGURE 3: TOTAL GRID IMPORT AND EXPORT OVER THE TRIAL PERIOD WITH AVERAGE MONTHLY EFFICIENCY IV FIGURE 4: RESULTS FROM THE MELINDA AVENUE FEEDER IN NEWCASTLE SHOW A REDUCTION IN PEAK DEMAND OF OVER 15% WITH THE USE OF REDFLOW'S ESS DURING TRADITIONAL PEAL TIMES V FIGURE 5: RESULTS FROM THE SCONE FEEDER IN MID-JUNE SHOW A TRANSFORMATION FROM PEAK TO TROUGH WITH THE USE OF REDFLOW'S ESS DURING TRADITIONAL PEAK TIMES V FIGURE 6: REDFLOW'S M90 90KW, 180KWH ESS WAS INSTALLED AT THE UNIVERSITY OF QUEENSLAND IN LATE MAY 2012 VI FIGURE 7: THE SGSC TRIAL IN NEWCASTLE HAS INTEGRATED A TOTAL OF 200KW OF SOLAR AND 50KW OF BLUE GEN FUEL CELL GENERATION WITH 40 REDFLOW ESS VI FIGURE 8: COMPONENTS OF REDFLOW'S R510 ESS 2 FIGURE 9: INSTALLATION OF AN ESS IN NEWCASTLE 2 FIGURE 10: THE LOCATIONS OF THE REDFLOW ESS IN NEWCASTLE, WITH THE FUEL CELL GENERATION CONNECTED TO EACH FEEDER 3 FIGURE 11: THE R510 ESS INSTALLED AT THE SMART HOME IN NEWINGTON, SYDNEY 4 FIGURE 12: TOTAL GRID IMPORT AND EXPORT OVER THE TRIAL PERIOD WITH AVERAGE MONTHLY EFFICIENCY 6 FIGURE 13: A TYPICAL FEED-IN CURVE FOR THE 4 ESS ON THE MELINDA AVENUE FEEDER 6 FIGURE 14: THE REDUCTION IN PEAK DEMAND SEEN BY THE GRID (4-10) IN MAY ON THE MELINDA AVENUE FEEDER 7 FIGURE 15: REDUCTION IN PEAK DEMAND SEEN BY THE MELINDA AVENUE FEEDER DURING TRADITIONAL PEAK TIMES 7 FIGURE 16: THE PEAKS CAUSED BY AUTOMATIC HOT WATER SYSTEMS ON THE MELINDA AVENUE FEEDER 8 FIGURE 17: AVAILABILITY OF ZBMS AND ESS IN NEWCASTLE 8 FIGURE 18: AVERAGE POWER ON THE SCONE RECLOSER 14 TO 18 MAY (BEFORE REDFLOW ESS COMMENCED FULL OPERATION) 9 FIGURE 19: AVERAGE POWER ON THE SCONE RECLOSER 18 TO 22 JUNE (WITH REDFLOW ESS OPERATIONAL) 9 FIGURE 20: AVAILABILITY OF ZBMS AND ESS IN SCONE 10 FIGURE 21: THE ESS CAN DYNAMICALLY FOLLOW THE LOAD, AND ANY SUDDEN CHANGES TO GENERATION 10 11/10/2012 Page viii

9 1 Introduction Increases in both general and peak electricity demand, the integration of intermittent and distributed generation, and developments in communications technology have all enabled, as well as necessitated a more advanced electricity network. This moves away from the traditional unidirectional grid, with central generation and one-way communications. As such, utilities have seen Smart Grids as an answer to the need for greater grid capacity, as well as allowing transmission and distribution infrastructure upgrade deferral. According to the Electric Power Research Institute (EPRI), a Smart Grid integrates and enhances other necessary elements including traditional upgrades and new grid technologies with renewable generation, storage, increased consumer participation, sensors, communications and computational ability [2]. The use of energy storage in Smart Grids can provide utilities with many benefits, including improved operational efficiency and increased value of distributed generation, thereby improving customer satisfaction [3]. This is done through the time-shifting of electricity from low demand, as well as times of high distributed generation output, to high demand times. However, despite the many benefits of energy storage, existing proven technologies are either inappropriate for Smart Grid applications (in the case of lead acid batteries), or are physically impossible to implement on a large scale in a variety of areas (in the case of pumped hydro or compressed air energy storage (CAES)). As such, new and advanced technologies are emerging to satisfy this new market. These must be trialed and tested to ensure that they are both reliable and align with the needs of utilities. As such, the Australian Government has initiated, and is currently funding the Smart Grid, Smart Cities (SGSC) Trial. It aims to test several hypotheses about many types of Smart Grid technology. In particular, the Trial aims to quantify the following seven benefits of energy storage: 8. Reduction in peak demand energy storage as a cost-effective and reliable alternative to network capacity expansion 9. Improvement in network reliability/voltage/power factor/power quality costeffectiveness and value of energy storage 10. Energy supply during peak price events net benefit to retail sector 11. Minimisation of customers energy bills with innovative tariffs e.g. time of use together with energy storage. 12. Combined benefit between consumer, retail and network sectors 13. Investigation of large capacity (~1MVA) storage extra cost and other benefits compared to smaller capacity storage 14. Intermittent generation support optimisation of renewable energy sources value RedFlow has been selected as one of several advanced energy storage companies to supply energy storage systems (ESS) for the trial to test these seven benefits of energy storage to Smart Grids. The ESS supplied for the Trial by RedFlow are the zinc-bromide module (ZBM)-based R510 model (see Appendix B R510 Product Brochure). These systems are rated at 5kW, 10kWh output, and comprise one ZBM, inverter, 3G modem for communications, battery management system (BMS), remote terminal unit (RTU) and other power electronics housed in a metal enclosure (see Figure 8). 11/10/2012 Page 1

10 Power Electronics Including BMS, RTU, 3G Modem Inverter SMA Sunny Backup Enclosure 1850 H x 971 L x 567 W ZBM Figure 8: Components of RedFlow's R510 ESS 61 R510 ESS have been installed (see Figure 9) and commissioned on separate residential properties as part of the SGSC Trial. This includes 40 systems in Newcastle (see Figure 10), which were installed throughout late 2011 and early All systems were operational by February 2012, and operate on four different charge/discharge cycles with the aim of reducing the peak load seen by the grid in these suburban areas. They are also integrated with solar and fuel cell distributed generation. Figure 9: Installation of an ESS in Newcastle 11/10/2012 Page 2

11 Figure 10: The locations of the RedFlow ESS in Newcastle, with the fuel cell generation connected to each feeder Twenty systems were installed in Scone (a map cannot be given for privacy reasons) in early 2012, and were operational by May These systems are located at fringe-of-grid areas, and are being used to test islanding aspects of a Smart Grid, as well as integration with wind turbines. 11/10/2012 Page 3

12 One R510 ESS (see Figure 11) has also been in operation since October 2011 in Ausgrid s Smart Home, located in the suburb of Newington in Western Sydney. It was installed as an upgrade to an existing RedFlow lead-acid system that was installed in mid-2010, with a ZBM installed in late 2010 to augment the lead acid storage. This ESS originally operated on a daily charge/discharge cycle, which was followed by a zero import/export from the grid mode with the installation of the R510. This has recently been modified to a hybrid cycle including both forced charge/discharge, as well as dynamic load following. Figure 11: The R510 ESS installed at the Smart Home in Newington, Sydney This Case Study provides an overview and analysis of the performance of RedFlow s R510 ESS in their Smart Grid applications. It will cover the period of 1 March to 21 May It has been divided into the sections of Operation, Results and Lessons Learnt. Each section addresses the three locations of Newcastle, Scone and Newington separately. Appendix A contains a list of abbreviations used in this Case Study. Appendix B contains a Product Brochure for RedFlow s R10, outlining key specifications of the ESS. Appendix C contains a list of all sites testing energy storage in the SGSC trial (see Appendix D for an explanation of schedule profiles). Appendix D contains a list of charge and discharge schedule profiles. Appendix E contains a large graph of peak reduction data from Newcastle. 11/10/2012 Page 4

13 2 Operation 2.1 Newcastle The 40 R510 ESS installed in Newcastle were located in suburban areas, which generally experiences a very high standard of power quality. However, in relation to the installation of further embedded generation (e.g. fuel cells and extra solar panels), the value and capabilities of energy storage in meeting these non-traditional challenges to the grid were tested. As such, the ESS were set to charge from and discharge into the grid at set times every weekday. The ESS did not operate at all (apart from discharging in the early hours of each Saturday) on weekends. This time schedule was varied between four different profiles: Profile A, Profile B, Profile C and Profile D (see Appendix D). 10 ESS operated on each profile (see Appendix C1). The efficiency and reliability of the R510 ESS in reducing the peak demand seen by the grid was tested. The results of this are presented in Section Scone The 20 R510 ESS installed in Scone were located in fringe of grid areas, which generally experiences a lesser standard of power quality than those areas in Newcastle. As such, the islanding abilities of the ESS were tested, as well as their ability to limit peaks seen by the grid. The ESS did not operate at all (apart from discharging in the early hours of each Saturday) on weekends. This time schedule was varied between two different profiles: Profile E and Profile F, (see Appendix D). 6 ESS operated on Profile E, and 14 ESS operated on Profile F (see Appendix C1). 2.3 Newington The R510 ESS installed in Newington is located in a suburban area, which generally experiences a very high standard of power quality. While the ESS at Newington has operated under several types of cycles, it is currently operating under Profile G (see Appendix D). When it is in the load following mode, the ESS discharges power to the household load if the generation from the solar panels and the fuel cell in insufficient, and charges from any excess generation that exceeds the household load. The efficiency and reliability of the R510 ESS in reducing the peak demand of the Smart Home seen by the grid was tested. The results of this are presented in Section /10/2012 Page 5

14 Grid Feed Power (kw) kwh % 3 Results Over the period between 1 March and 31 May 2012, the R510 ESS installed for the SGSC trial have outputted a total of MWh to the grid at an average efficiency of 58.03%. This is shown in the graph below in Figure SGSC ESS Grid Interaction March April May Grid Import (kwh) Grid Export (kwh) Efficiency (%) 0 Figure 12: Total grid import and export over the trial period with average monthly efficiency 3.1 Newcastle Peak Demand Reduction The ESS installed in Newcastle have had significant impact on the peak demand seen by the grid. Analysis has shown that during the cooler Autumn months, a ratio of 4 ESS to 65 residential customers produced reductions of peak demand seen by the grid during the traditional peak period (between the hours of 4pm and 10pm on weeknights) by an average of 5.39% in May. This is with 4 ESS operational for most of the month (1 ESS recommenced operation on 10 May after being out of service since the start of the month), and therefore a ratio of approximately 1 ESS for 16 customers. A typical output curve of the cumulative ESS is shown in Figure ESS on Profile A feed in 9kW total Power Fed into the Grid by 4 ESS 3 ESS on Profile A are almost flat 1 ESS on Profile B feeds in 3kW 4:00 5:00 6:00 7:00 8:00 9:00 10:00 Figure 13: A typical feed-in curve for the 4 ESS on the Melinda Avenue feeder 11/10/2012 Page 6

15 Grid Power (kw) Reduction in Peak Demand (%) However, the irregularity of this curve results in only very small peak reductions when peaks occur around 7pm, or after 9pm. The variability in peak reduction can be seen in Figure Reduction in Peak Demand (4pm to 10pm) in May Peak occurs at times of low ESS feed-in power Figure 14: The reduction in peak demand seen by the grid (4pm-10pm) in May on the Melinda Avenue feeder Despite this, the ESS have been capable of achieving peak reduction of about 15% (see Figure 15). A grid peak demand graph for the week of Monday 9 April 2012, showing similar results, is given in Appendix E. Grid-Feed Period Monday 9 April Peak Reduction 15.19% 65 Residential Customers 4 R510 ESS :40 4:10 4:40 5:10 5:40 6:10 6:40 7:10 7:40 8:10 8:40 9:10 9:40 10:10 Real Power (kw) with Storage Real Power (kw) without Storage Figure 15: Reduction in peak demand seen by the Melinda Avenue feeder during traditional peak times However, in this particular area of Newcastle, many homes are fitted with automatic hot water systems that are set to charge at 11pm. As can be seen in Figure 16, this causes a much larger peak than that seen during traditional peak periods, when the ESS are discharging into the grid. Therefore, while the ESS are effectively reducing the traditional evening peak, they are not addressing the actual peaks see by the grid. 11/10/2012 Page 7

16 % kw Effect of Hot Water Systems Peak Reduction shown in Figure 15 Monday 9 April 43.29% increase of traditional evening peak 65 Residential Customers 4 R510 ESS Tuesday 10 April 0 12:00 AM 6:00 AM 12:00 6:00 12:00 AM 6:00 AM 12:00 6:00 12:00 AM Real Power (kw) with Storage Real Power (kw) without Storage Figure 16: The peaks caused by automatic hot water systems on the Melinda Avenue feeder This highlights the importance of thorough load profiling prior to setting charge/discharge profiles for residential ESS aimed at reducing peaks in demand seen by the grid. In the case of the Newcastle ESS involved in the SGSC trial, this would have raised two main issues. Firstly, the ESS should be discharging between about 11pm and 2am to reduce the true peak caused by the hot water systems. Furthermore, the discharge profiles should be more appropriately staggered to achieve a smoother collective output curve, that discharges more power during times when the load curve is normally highest. These measures would aid in reducing the peak demand seen by the grid even more than the significant reductions already provided by the ESS Reliability As the ZBM is still an emerging technology, the reliability of the R510 is an important aspect of evaluating its value in Smart Grid applications of energy storage. As can be seen in Figure 17, the vast majority of ZBMs were available for operation throughout the study period, showing that ZBM failure is not the main cause for making ESS unavailable Percentage of Newcastle ZBM and ESS Available Mar 13-Mar 25-Mar 6-Apr 18-Apr 30-Apr 12-May 24-May Newcastle ZBM Available Newcastle ESS Available Figure 17: Availability of ZBMs and ESS in Newcastle 11/10/2012 Page 8

17 Power (kva) Power (kva) 3.2 Scone Peak Demand Reduction The ESS installed in Scone have had significant impact on the peak demand seen by the grid. This is due in part to the high ratio of ESS to customers (just over 1 ESS per customer) in this semi-rural area. Analysis has shown that during the cooler Autumn and Winter months, the usual evening peak seen by the grid (see Figure 18) is transformed into a noticeable trough with the use of RedFlow s ESS (see Figure 19). Power on Miranee Road Recloser No Energy Storage in Use 13 Residential Customers :00 AM 3:00 AM 6:00 AM 9:00 AM 12:00 3:00 6:00 9:00 12:00 AM Figure 18: Average power on the Scone recloser 14 to 18 May (before RedFlow ESS commenced full operation) Power on Miranee Road Recloser Energy Storage in Use Grid Charge Period 28kW Grid Charge Period 12kW Grid- Feed Period 45kW 13 Residential Customers 15 R510 ESS 12:00 AM 3:00 AM 6:00 AM 9:00 AM 12:00 3:00 6:00 9:00 12:00 AM Figure 19: Average power on the Scone recloser 18 to 22 June (with RedFlow ESS operational) However, it can also be seen that peaks occur very late at night, and in the early morning due to automatic off-peak hot water systems turning on. However, these hot water system peaks are not significantly larger in magnitude than the traditional morning and evening peaks. 11/10/2012 Page 9

18 Power (kw) % Reliability As the ZBM is still an emerging technology, the reliability of the R510 is an important aspect of evaluating its value in Smart Grid and islanding applications of energy storage. As can be seen in Figure 20, the vast majority of ZBMs were available for operation throughout the study period, showing that ZBM failure is not the main cause for making ESS unavailable Percentage of Scone ZBM and ESS Available Apr 11-Apr 18-Apr 25-Apr 2-May 9-May 16-May 23-May 30-May Scone ZBM Available Scone ESS Available Figure 20: Availability of ZBMs and ESS in Scone 3.3 Newington Load Following Results, such as those presented in Figure 21, show that the R510 ESS can charge and discharge according to the needs of the load, as well as discharge when interruptions to normal generation occur. This is known as load following. Blue Gen Fuel Cell Outage Performance :00 AM 2:00 AM 4:00 AM 6:00 AM 8:00 AM 10:00 AM 12:00 Fuel cell outage 2:00 4:00 ZBM discharges to satisfy load ZBM discharges in preparation to strip 6:00 8:00 InvPwrAt [kw] Blue Gen Power Total Usage 10:00 Figure 21: The ESS can dynamically follow the load, and any sudden changes to generation 11/10/2012 Page 10

19 3.3.2 Reliability The R510 at Newington was installed on 5 October It engaged in daily operation until 20 December, when it was shut down because its ZBM s health was deteriorating and a fault was expected to soon occur. Soon after, the resident family moved out of the Smart Home and Ausgrid suspended the operation of most of the Smart Home s elements while it underwent upgrades and maintenance. The R510 ESS was kept shut down until late March when the ZBM was replaced. After a series of calibration and test cycles, the ESS was shut down again until the new resident family moved into the Smart Home on 28 April, at which point the ESS recommenced operation. It has not experienced a fault since this time. At the end of May 2012, the new ZBM installed at the Smart Home had undertaken 48 cycles (of less than 100% depth of discharge due to the shortage of excess generation from the Smart Home to charge the ZBM), and outputted 87.23kWh. 11/10/2012 Page 11

20 4 Lessons Learnt 4.1 Newcastle The 40 systems in Newcastle were the first R510s to be installed on a large scale for RedFlow. As a result of this, and early operation, the following important points were observed for future work in the area. The use of specialist installers, O Donnell Griffin (ODG), has been highly beneficial in the rollout of ESS to Newcastle. This brought expertise to the project that RedFlow did not have, and will be used in future large-scale trials. Small issues with Newcastle installations shaped changes to the process during Scone installations. RedFlow needs to have unlimited access to data sent from each ESS. In the case of the SGSC project, Ausgrid kept data on their machines, which made it highly difficult and timeconsuming for RedFlow to acquire and analyse data from systems for both monitoring and improvement purposes. An average of 76.14% of the ESS installed for the SGSC Trial were available for operation over this study period. While this can be improved, the average of 96.65% availability of ZBMs over the same time period means that most ESS faults were not due to ZBM failures. Instead, it was elements of the ESS, and in particular the analogue looms that caused most ESS faults. This shows the reliability of the core ZBM product, as well as the need for specialist system integrators to be involved in the design and manufacturing of future ESS. It is important to carry out appropriate load and distributed generation analyses prior to project commencement. In the case of Newcastle, this would have shown the large peaks caused by the automatic hot water systems. In response to this, and in addressing the true peaks in the Newcastle area, the discharge period of the ESS should be shifted, or hot water systems should begin charging at staggered times. The ESS have been installed at suitable ratios (1 ESS for every 16 customers) to the number of customers in respective areas to achieve reductions in peak demand seen by the grid by about 5-10%. However, reductions could be improved with different discharge profiles. Regardless, the number of ESS may need to be increased to see the same peak reduction results during the summer months. 4.2 Scone The following lessons were also noted for the installation and operation of the 20 ESS in Scone. The use of specialist installers, O Donnell Griffin (ODG), has also been highly beneficial in the roll-out of ESS to Scone. RedFlow needs to have unlimited access to data sent from each ESS. In the case of the SGSC project, Ausgrid kept data on their machines, which made it highly difficult and timeconsuming for RedFlow to acquire and analyse data from systems for both monitoring and improvement purposes. 11/10/2012 Page 12

21 An average of 79.57% of the ESS installed for the SGSC Trial were available for operation over this study period. While this can be improved, the average of 92.24% availability of ZBMs over the same time period means that most ESS faults were not due to ZBM failures. This shows the reliability of the core ZBM product, as well as the need for specialist system integrators to be involved in the design and manufacturing of future ESS. 4.3 Newington The R510 ESS, as well as previous ESS, installed at the Newington Smart Home have produced the following lessons learnt from installation and operation. The 5kW, 10kWh rating of the R510 ESS is appropriate for the needs of the Smart Home household load (approximately 15kWh/day), and would still be suitable for slightly higher loads. Large and continuous loads, such as the EV, are often too large for the ESS to sustain without requiring power from the grid. There is a need for a fuel cell or other similar reliable form of embedded generation to backup solar generation in off-grid or minimum grid import Smart Grid applications. The effectiveness of the R510 in conjunction with embedded generation to load follow and greatly reduce grid import to below 8% of the time. 11/10/2012 Page 13

22 5 Conclusions Overall, this project has shown that ZBMs in R510 ESS are a suitable type of energy storage to use in Smart Grid applications. The ZBM is reliable and effective in reducing peak demand seen by the grid, as well as in islanding applications. Overall, RedFlow has learnt many valuable lessons over the course of the SGSC Trial, and will use these to improve upon their technology and installation procedures for future Smart Grid ESS designs. These have included: The benefits of using specialist installers to carry out large-scale installations of ESS. The importance of complete access to trial data for monitoring and analysis. The ZBM is a reliable product, and was by far not the main cause of ESS failures. The subsequent need for specialist system integrators to be involved in the design and manufacturing of future ESS. 11/10/2012 Page 14

23 6 References [1] Ausgrid. (2011) Program Trial Areas, Ausgrid. [Online]. Available: City/~/media/Microsites/SGSC/Files/PDFs/Smart%20Grid%20Smart%20City%20project%20trial %20map.pdf [2] Electrical Power Research Institute (EPRI), Estimating the Costs and Benefits of the Smart Grid: A Preliminary Estimate of the Investment Requirements and the Resultant Benefits of a Fully Functioning Smart Grid, EPRI, Palo Alto, CA, Final Report , Mar [3] M. Dean. (2012, Jun. 20). Smart Grid growth to spur demand for energy storage [Online]. Available: id/2012/june/-smart_grid_growth.html 11/10/2012 Page 15

24 Appendix A List of Abbreviations BMS - Battery Management System CAES - Compressed Air Energy Storage EPRI - Electric Power Research Institute ESS - Energy Storage System RTU - Remote Terminal Unit SGSC - Smart Grid, Smart Cities ZBM - Zinc Bromide Module 11/10/2012 Page 16

25 Appendix B R510 Product Brochure 11/10/2012 Page 17

26 11/10/2012 Page 18

27 11/10/2012 Page 19

28 Appendix C List of Trial Sites 11/10/2012 Page 20

29 Appendix C1 List of Newcastle Sites Site ZBM # Date of Cabinet ID Number (Class) Commissioning Schedule Status Comments R510-A1-DE November 2011 Profile A Operational Leak 1 Trip R510-A1-DE November 2011 Profile A Operational (A+) R510-A1-DE November 2011 Profile A Operational 2 ZBM Replacements R510-A1-DE November 2011 Profile A Not Operational (A+) R510-A1-DE November 2011 Profile A Operational ZBM Replacement (A+) R510-A1-DE November 2011 Profile A Operational ZBM Replacement Previous SMA Fault (A+) R510-A1-DE November 2011 Profile A Unknown Fault ZBM Replacement Q4 Switching (A+) R510-A1-DE November 2011 Profile A Not Operational System Noise ZBM Replacement (A+) R510-A1-DE November 2011 Profile A Operational ZBM Replacement (A+) R510-A1-DE November 2011 Profile A Operational Analog Loom Fixed ZBM Replacement (A+) R510-A1-DE November 2011 Profile B Not Operational System Noise ZBM Replacement (A+) R510-A1-DE November 2011 Profile B Operational ZBM Replacement (A+) R510-A1-DE November 2011 Profile B Not Operational ZBM replacement (A+) R510-A1-DE November 2011 Profile B Operational ZBM Replacement (A+) R510-A1-DE November 2011 Profile B Operational ZBM Replacement (B) R510-A1-DF December 2011 Profile B Operational (A+) R510-A1-DF December 2011 Profile B Operational ZBM Replacement (A) R510-A1-DF December 2011 Profile B Operational Leak1 Trip and Battery Voltage Fail (A) R510-A1-DF December 2011 Profile B Not Operational BC Firmware Upgrade Required (A) R510-A1-DF December 2011 Profile B Operational Incorrect Installation (A+) R510-A1-DF December 2011 Profile C Not Operational Previous RTU Fault ZBM Replacement 11/10/2012 Page 21

30 Site ZBM # Date of Cabinet ID Number (Class) Commissioning Schedule Status Comments (A) R510-A1-DF December 2011 Profile C Not Operational System Noise (A) R510-A1-DE December 2011 Profile C Operational Previous Analog Loom Failure (A) R510-A1-DF December 2011 Profile C Operational Incorrect Installation (A) R510-A1-DF December 2011 Profile C Not Operational (A+) R510-A1-DF December 2011 Profile C Operational Incorrect Installation ZBM Replacement Previous Analog Loom Fault (A) R510-A1-DF December 2011 Profile C ZBM Fault RTU Lost Comms with BC (A+) R510-A1-DE December 2011 Profile C Operational ZBM Replacement (A) R510-A1-DF December 2011 Profile C Operational (B) R510-A1-DF December 2011 Profile C Operational Previous Analogue Loom Fault (A) R510-A1-DG December 2011 Profile D Not Operational (A+) R510-A1-DG December 2011 Profile D Operational ZBM Replacement (A) R510-A1-DG December 2011 Profile D Operational (A+) R510-A1-DG December 2011 Profile D Not Operational ZBM Replacement (A) R510-A1-DG December 2011 Profile D Unknown Fault Mains Fail, Needs AS Board Replacement (A) R510-A1-DG December 2011 Profile D Operational (A+) R510-A1-DG December 2011 Profile D Operational ZBM Replacement (A+) R510-A1-DG December 2011 Profile D Battery Controller Amp Lockout Failure Fault ZBM Replacement (A) R510-A1-DG December 2011 Profile D Operational (A+) R510-A1-DG December 2011 Profile D Operational ZBM Replacement 11/10/2012 Page 22

31 Appendix C2 List of Scone Sites Site ZBM # Number (Class) Cabinet ID Date of Commissioning Schedule Status Comments (A+) R510-A1-DG March 2012 Profile F Operational (A+) R510-A1-DG March 2012 Profile E Operational ZBM Replacement (A+) R510-A1-DG March 2012 Profile F Operational ZBM Replacement (A+) R510-A1-DG April 2012 Profile F Operational (A+) R510-A1-DH March 2012 Profile F Operational ZBM Replacement (A+) R510-A1-DH March 2012 Profile E Operational (A+) R510-A1-DH April 2012 Profile F Operational (A) R510-A1-DH April 2012 Profile F Operational Previous Analog Loom Fault (A+) R510-A1-DH March 2012 Profile E Operational ZBM Replacement (A+) R510-A1-DH April 2012 Profile F Operational (A+) R510-A1-DH March 2012 Profile E Operational ZBM Replacement (A) R510-A1-DH March 2012 Profile F Operational Previous ZBM Fault (A+) R510-A1-DH March 2012 Profile F Operational ZBM Replacement (A) R510-A1-DH March 2012 Profile F Operational Previous Battery Controller Fault (A+) R510-A1-DH March 2012 Profile F Operational ZBM Replacement (A) R510-A1-DH March 2012 Profile E Operational (A+) R510-A1-DH March 2012 Profile F Not Operational ZBM Replacement Grid Fault (A) R510-A1-DH March 2012 Profile F Operational Previous Battery Controller Fault (A) R510-A1-DH March 2012 Profile E Operational (A) R510-A1-DH March 2012 Profile F Operational 11/10/2012 Page 23

32 Appendix D Cycle Profiles Profile A Profile B Profile C Profile D Profile E Profile F Profile G Profile Days of the Week Active Time Power Charge 2am to 8am 2kW Monday to Friday Discharge 4pm to 2am 3kW Charge 2am to 8am 2kW Monday to Friday Discharge 7pm to 2am 3kW Charge 2am to 8am 2kW Monday to Friday Discharge 5pm to 2am 5kW Charge 2am to 8am 2kW Monday to Friday Discharge 7pm to 2am 5kW Charge 8am to 2pm 2kW Monday to Friday Discharge 5pm to 2am 3kW Charge 1am to 7am 2kW Monday to Friday Discharge 6pm to 1am 3kW Charge 2am to 7am 2kW Monday, Wednesday, Friday, Load Follow 7am to 11pm Load Following Sunday Discharge 11pm to 2am 4.5kW 11/10/2012 Page 24

33 12:00 AM 4:00 AM 8:00 AM 12:00 4:00 8:00 12:00 AM 4:00 AM 8:00 AM 12:00 4:00 8:00 12:00 AM 4:00 AM 8:00 AM 12:00 4:00 8:00 12:00 AM 4:00 AM 8:00 AM 12:00 4:00 8:00 12:00 AM 4:00 AM 8:00 AM 12:00 4:00 8:00 12:00 AM 4:00 AM 8:00 AM 12:00 4:00 8:00 12:00 AM 4:00 AM 8:00 AM 12:00 4:00 8:00 12:00 AM 4:00 AM 8:00 AM Grid Power (kw) Appendix E Newcastle Reduction in Peak Demand Effect of Storage on Peak Reduction - April Friday 6 April Saturday 7 April Sunday 8 April Monday 9 April Tuesday 10 April Wednesday 11 April Thursday 12 April Friday 13 April Peak Reduction 17.78% Peak Reduction 15.19% Peak Reduction 14.78% Peak Reduction 14.13% Peak Reduction 14.85% Real Power (kw) with Storage Real Power (kw) without Storage 65 Residential Customers 4 R510 ESS Operating 11/10/2012 Page 25