Decentralized Battery Energy Management for Stand-Alone PV- Battery Systems Umarin Sangpanich (PhD.) Faculty of Engineering at Sriracha Kasetsart University (Sriracha campus) 19 May 2016
Outline A key of stand-alone renewable energy systems Battery management A traditional battery control method Decentralized battery storage control methods Decentralized Battery Energy Management (DBEM) method Optimization of a stand-alone PV-Battery system with the DBEM method Case study Results and Discussions Conclusions Prototype Publication 2
A key of stand-alone renewable energy systems Battery management A traditional battery control method is a single charge controller by grouping all batteries into parallel strings. Photovoltaic array Control and Metering system Charge controller DC bus AC/DC converter Inverter AC wind turbines Dispatched load Drawbacks: Batteries in the same string must have the same capacity and initial State of Charge (SOC). If a battery has a higher initial SOC, it would be charged faster and have gassing while other batteries in the string are still undercharged, leading to the reduction of battery lifetime. Due to limitation of charging power levels, therefore, the energy produced will be wasted and the wasted energy will increase in larger battery energy storage systems. 3
Charging process situation PV output power Waste energy Cloudy day Waste energy Waste energy One large group Time Minimum current limit of batteries 4
A key of stand-alone renewable energy systems Battery management Decentralized Photovoltaic (PV) and battery system with multilevel inverter -Better performance -Longer lifetime of battery storage systems However, larger systems will be more expensive. Decentralized battery storage control system Source : M. I. Desconzi etc., 2010 PV array DC-DC S1 S2 S3 S4 S5 S6 Inverter Load Source: R. Kaiser, 2007 5
Decentralized Battery Energy Management (DBEM) method PV array Control center Charger Inverter Load Objectives: To minimize. loss of power supply cost of energy wasted electrical energy To prolong battery life Multi-switch groups DBEM system Different group sizes **The system should be practical and economical for operation.*** 6
Charging process situations Decentralized Battery Energy Management (DBEM) method Sunny day Cloudy day PV output power Small group Large group Small group PV output power Large/Medium groups Small group Small group Medium group Time Medium group Time 7
Charging and discharging processes of the DBEM method P pv, P load P pv >P load Yes I min,rated I pv, remain I max,rated Yes Charge a battery group having the lowest voltage No Discharge a battery group having the highest voltage (V 1 ) No Charge a battery group having higher voltage by checking rated current Check V batt & I rated in each group Charging process Discharge the battery group until V 1 =V 2 if P pv <P load Discharge the battery groups until V 1 =V 2 =V 3 if P pv <P load Discharging process P pv is PV power. P load is load power. V batt is battery voltage. I pv,remained is PV current remained from supplying load. I min,rated and I max,rated are minimum and maximum rated battery current of each battery group, respectively. 8
Optimization of a stand-alone PV-Battery system with the DBEM method Load demand Solar radiation & Air temperature data SPEA2 operation No Random initial populations: number of PVs, batteries Operation of a PV-Battery system: calculate LPSP, LCOE Maximum number of generation reached Objectives: To minimize Levelized Cost of Energy (LCOE) To minimize Loss of Power Supply Probability (LPSP), 0 LPSP 1. New populations Yes Final population: N pv & N batt Optimal solution: minlpsp, minlcoe Optimization method: Strength Pareto Evolutionary Algorithm 2 (SPEA2) 9
Solar radiation (kw/m2) Case study Average hourly solar radiation in Thailand 1 A load profile in a Thai rural area 1.4 1.2 0.8 1 0.8 0.6 0.6 0.4 0.2 0.4 0 400 300 200 Day 0.2 25 20 15 100 10 5 Hour 0 0 0 Estimated costs and lifetime of system components Components Initial capital cost Replacement cost O&M cost Lifetime PV module including Tax 15,000 Baht per module - Installation cost and other components 20% of total PV system cost 15,000 Baht per module 0.5% of ICC 25 years Battery, 550 Ah 12,500 Baht per cell 12,500 Baht per cell 3% of ICC 1000 cycles Bi-directional converter and system control 22,550 Baht per kw 22,550 Baht per kw 3% of ICC 15 years Charge controller 190 Baht per Ampere 190 Baht per Ampere 3% of ICC 15 years 10
Results and Discussions LPSP and LCOE of stand-alone PV-Battery systems The fractions of the three groups were set as 1/6, 1/3 and 1/2 to be small, medium and large groups, respectively. 11
Results and Discussions Comparisons between one and three battery groups Particular Battery groups 1 group 3 groups Loss of power supply probability, LPSP 0% 1% 32.11% 1% 0% ( 87 hrs) (2,813 hrs) ( 87 hrs) PV capacities (kwp) 35 31 20.5 20.5 13 Battery capacities (kwh) 356.4 237.6 316.8 316.8 237.6 Annual PV energy (MWh/year) 67.62 59.89 43.5 43.5 25.12 Annual total waste energy (MWh/year) 25.33 17.6 1.40 1.21 0.32 Annual overall energy efficiency of systems 62.5% - - 97.22% - LCOE (Baht/kWh) 10.86 9.80 9.64 9.35 6.81 The 3-groups system has lower PV and battery capacities and lower annual total wasted energy, leading to higher overall energy efficiency and lower LCOE. The 3-groups system has higher reliability than the 1-group system. 12
02 AM, 30-Sep 07 AM, 30-Sep 12 PM, 30-Sep 04 PM, 30-Sep 09 PM, 30-Sep 02 AM, 1-Oct 07 AM, 1-Oct 12 PM, 1-Oct 04 PM, 1-Oct 09 PM, 1-Oct 02 AM, 2-Oct 07 AM, 2-Oct 12 PM, 2-Oct 04 PM, 2-Oct 09 PM, 2-Oct 02 AM, 3-Oct 07 AM, 3-Oct 12 PM, 3-Oct 04 PM, 3-Oct 09 PM, 3-Oct Power (kw) State of Charge, SOC Results and Discussions The SOC and PV power supply for load demand, of the system using one battery group 35 30 25 20 15 10 5 0 PV power Load power SOC 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Time, Date Cloudy days in rainy season 13
02 AM, 30-Sep 07 AM, 30-Sep 12 PM, 30-Sep 04 PM, 30-Sep 09 PM, 30-Sep 02 AM, 1-Oct 07 AM, 1-Oct 12 PM, 1-Oct 04 PM, 1-Oct 09 PM, 1-Oct 02 AM, 2-Oct 07 AM, 2-Oct 12 PM, 2-Oct 04 PM, 2-Oct 09 PM, 2-Oct 02 AM, 3-Oct 07 AM, 3-Oct 12 PM, 3-Oct 04 PM, 3-Oct 09 PM, 3-Oct Power (kw) State of Charge, SOC Results and Discussions 35 30 25 20 15 10 5 0 The SOC and PV power supply for load demand, of the system using three battery groups PV power Load power SOC of a small group SOC of a medium group SOC of a large group 1.2 1 0.8 0.6 0.4 0.2 0 Time/Date Cloudy days in rainy season 14
Conclusions The DBEM method is proposed for minimizing loss of power supply, cost of energy and wasted electrical energy. From SPEA2 optimization and energy simulation, the PV system using the DBEM method has higher reliability and energy efficiency than the system using one battery group, while decreasing the number of PV and battery modules leading to lower LCOE and waste energy. The DBEM method can be applied for renewable energy systems such as wind turbines. A DBEM prototype, of which controller and circuit are uncomplicated, has being designed and built. 15
Prototype System Specification: For a PV system of 3 kwp Using a micro processor to be a center controller between a charger and an inverter Data monitoring via Ethernet Upload data via USB Sleep mode for saving energy PV array Control center Charger Inverter Load DBEM system 16
Publications U. Sangpanich, Optimization of Photovoltaic Systems Using Batteries for Peak Demand to Improve Rural Electrification, CIGRÉ Canada Conference 2015 Proceeding, Canada, August 31-September 2, 2015 U. Sangpanich, A Novel Method of Decentralized Battery Energy Management for Stand-Alone PV-Battery Systems, in The 6 th IEEE PES Asia-Pacific Power and Energy Engineering Conference (IEEE PES APPEEC 2014), 2014. 17
Thank you very much for your attention. AC/DC converter Photovoltaic array AC/DC converter AC wind turbines AC wind turbines Control and Metering system Dispatched load Charge controller Inverter DC bus 18