Key Learnings from The Establishment of a Battery Energy Storage Testing Facility by Peter Langley, Eskom Research
Demand Demand Traditional vs Evolving Grid Peak Load Intermediate Load Base Load Time of Day Forecasted Demand, incl. solar and wind - 2035 Evolving Grid with inclusion of IPPs (based on IRP2010 and not IRP2016) Charge Discharge Discharge Forecasted Demand Less Solar Less Solar & Wind 2 Time of Day
Current Energy Storage Installations Palmiet Pumped Storage Installed capacity: 400MW (for 28 hrs) Ingula Pumped Storage Installed capacity: 1332MW (for 14 hrs) Drakensberg Pumped Storage Installed capacity: 1000MW (for 28 hrs) 3
Energy Storage as a Flexibility Tool Pumped storage can only be installed at discrete locations with suitable geography. Eskom has been looking at alternative means to store energy for the last 15 years, resulting in the formation of the Large Scale Energy Storage Research Portfolio, some 10 years ago. Technologies considered range from mechanical storage to batteries, chemical storage, thermal energy storage and super capacitors. Batteries are considered the most likely alternative means of storage in the future. A network of large batteries could be the solution to Eskom s needs for increased flexibility. Eskom must determine the REAL performance of these batteries prior to widespread installation, especially in view of the current costs. Comparative testing is therefore essential. Battery Energy Storage Testing Facility Rosherville 4
Impact and Value of Energy Storage Transfer of the available energy during off-peak periods to the high demand periods Storage charged from base load generating plant Generation profile with storage Generation profile without storage Storage used to maintain frequency and voltage by balancing supply and demand Peak demand for power supplied by peaking plant, running only a few hours each day Storage discharged into network Storage charged from base load generating plant Generation profile without storage 5
Battery Energy Storage Testing Facility Objectives: 1.Demonstrate the effectiveness of battery energy storage at a grid scale. 2.Test individual battery technologies under real operating test regimes 3.Identify the best technology for various applications 6 Specification: Capability to test 5 x 200kW 1.2MWh different battery technology units All tested under identical load discharge profiles Output of the 5 units is then synchronised and fed back into the grid, as if they were a single 1MW battery unit, in order to demonstrate the effectiveness of this form of energy storage. 4.Establish the probable life cycle of each of the various technologies under real working conditions 5.Establish the round trip efficiencies of the various units 6.Give Eskom insight into the future installation of commercial battery storage units of the Megawatt scale.
New Technologies for testing in 2018 Vanadium Redox Flow Battery Zinc Bromine Flow Battery 7
Battery Energy Storage Testing Facility Open Tender Process followed for the supply, delivery, installation and maintenance of systems. Unit 1 is a lithium ion phosphate battery supplied by BYD of China Unit 2 is a sodium nickel chloride battery supplied by General Electric of the USA In the process of negotiating to install 2 further flow battery technologies, that will be tested on behalf of their suppliers, as an independent third party test facility. There is space for one further unit, although additional units could be added at an adjacent area to the test facility at a later date. 8
Test Profiles The test protocols over 90-day test basis include: Load Shifting - 6 hours continuous output at 200kW per battery, off-peak charging available for 8 hours per day, each day for the 90 period. PV Cloud Compensation A second by second survey of the output of a 400kWp solar system during an intermittently cloudy day, plus load shifting will be repeated for 90 days. Wind smoothing - A typical wind farm daily profile established from a South African wind facility, (under afternoon storm conditions) which is used to supply the battery and/or grid, on a daily basis for 90 continuous days, with the battery expected to absorb and discharge to smooth the output. Power Quality - The battery operates at the top end of its charge and smooth out frequency and voltage changes resulting from demand changes. Other The facility is a tool for Eskom to simulate actual conditions and to complete predictive testing prior to selection, design and installation of battery energy storage systems on the network. 9
Test Results Test 1 successfully demonstrated that Load Shifting using energy storage can be smoothly achieved and that the Eskom base load units could be more efficiently operated to minimise costs and emissions, by introducing widespread energy storage. Test 2 has shown that an energy storage system, combined with a solar system can provide despatchable power, even under the harshest solar performance. The profile replicates (and exaggerates) an intermittently cloudy day with high ramp rates and solar output changing from 100% to 25% in a few seconds. The storage system recognises these changes and compensates for the output, either by providing additional energy to the grid, or by absorbing excess production. Furthermore the storage system absorbs energy during the night time off-peak period, and thereby allows the solar system to provide despatchable energy (at a predetermined level) from 06.30am to 21.00pm. One drawback of the Test 2 demonstration is that under real conditions the daily set-point will need to be determined prior to the day s known solar output 10
Test Results Test 3: Using the Test Facility to simulate installations Capex deferral applications on constrained networks Using real data collected from an identified constrained network By doing simulations at the test facility it was successfully demonstrated that energy storage can be used for capex deferral. By simulating the use of energy storage and PV the constraint was removed on the line as well as allowing for the connection of new customers waiting to be connected to the grid. 11
Conclusion Battery energy storage is effective at grid scale It has been possible to test 2 different technologies under directly competitive conditions that replicate the probable operational duty cycles that battery energy storage will be required to perform in South Africa. Further testing will be carried out to establish the longevity of these technologies. A great need identified to expand the number of units under test and to have different technologies for evaluation. The identification of which technologies are best suited to which applications is an on-going exercise, although the results to date suggest that the one system under test is not suitable for use within Eskom. The batteries will need to be used to the end of their useful lives before generic estimates of life can be made. However, indicative performances at the end of the first 3 year test period can be estimated. 12
Conclusion (cont.) The project has successfully established round trip efficiencies for both of the technologies. These efficiencies differ under different operating cycles. The lithium ion system has an operating round trip efficiency of between 78 and 81%, whilst the other system is between 65 and 72%. It is believed that the lessons learnt during this research project would allow Eskom to go out on commercial tender for energy storage devices and to successfully integrate these devices into the grid. 13
Questions