24/05/2019

1

© 2019, all rights reserved.

1st Grid Transformation Professional Development Seminar on

Microgrid Design & Battery Storage

16:00 pm Battery operation in a microgrid A/Prof. Nesimi ERTUGRUL

Electrical System: General Configuration and Specifications

• 3x LG Chem R800 racks (total 273 kWh).

• Expandable to 364 kWh.

• 3x ABB PCS100 inverter modules (total 270 kVA).

• Expandable to 360 kVA.

• 350 kVA isolating transformer.

• SEL700GT grid-tie relay.

• SEL751A feeder relay.

• Flexible connection options.

24/05/2019

2

Slide

The internal battery is LG CHEM 273kWh Li-Ion battery, comprised of three racks of battery modules, located in a dedicated battery room of the shelter

AESKB System Description

Battery Operation in the Microgrid System:Preliminary Battery Tests

0

20

40

60

80

100

120

140

80% 85% 90% 95% 100%

Ch

arge

Rat

e (k

W)

Battery SoC

Transition of Constant Power ChargingInitial charge rate, 129kW

Initial charge rate, 80kW

Normal Max SoC

640

650

660

670

680

690

700

710

720

730

0%10%20%30%40%50%60%

DC

Vo

ltag

e o

f th

e B

atte

ry (V

)

Battery SoC

Discharging Test

Change to 16kW discharge / battery at 28 deg C / all module fans on (648V)

Turned-off at "Over-Discharge Power" 16kW at 1% (646V)

PaDEC system stopped discharge at 2% SoC (651V)

Change to 32kW discharge / battery at 28 deg C / all module fans on (655V)

798

800

802

804

806

808

810

812

814

816

818

88% 90% 92% 94% 96% 98% 100%

DC

Vo

ltag

e o

f th

e B

atte

ry (V

)

Battery SoC

Charging Test

Inverter shut down automatically at 99% - battery volts then settled to 815.9V dc

The system was stepped up to 240kW (92%) power (red) and reached 84% state of charge (SoC, blue), while charging from the Thebarton Diesel Generator.

24/05/2019

3

• The figure shows the DC power into the battery (calculated from nodes B and D in the system diagram).

• At the start of the inverter was told to draw zero power from the grid. Because the inverter requires power to operate, this was taken from the battery and results in short dips of negative battery power.

• Soon after the inverter was told to draw power from the grid to charge the battery, with power level increasing in steps up to 20kW.

• The actual DC power delivered to the battery was less than the 20kW (~18kW) the inverter consumes from the grid, because the inverter is not 100% efficient and includes some losses.

• The battery voltage (measured at node C) rises at an approximately constant rate for the duration of the constant power charging.

• There is a small voltage drop once the charging starts and the inverter starts to load the grid connection.

Operation of the system for a 20kW charge test at Thebarton

Inverter efficiency during the 20kW charge test.

5 10 15 20 25PDC0

20

40

60

80

100

Inve

rte r

Effi

c ien

c y(%

)

(kW)

The full output power of the inverter and battery tests

• The test starts by charging the battery for 15 minutes to just over 91% State of charge.

• The full power discharge test starts just after 17:30 by turning the 100kW load bank on. Initially, this only consumes power from the grid.

• After the load bank was running, the inverter was then enabled at run just under full power. The inverter outputs 258kW, with 113kW sent to the load and 143kW exported to the grid. The missing 2kW represents losses in the isolation transformer between the inverter and both the grid and load connections.

• The peak DC power (calculated from nodes B and D) was measured at -264kW, representing power flowing out of the battery. Because of inverter and transformer losses, not all battery power is delivered to the grid.

• The battery power that doesn’t reach the grid and load connections is lost as heat and must be removed by the air conditioning systems for the battery, inverter and transformer.

24/05/2019

4

Battery voltage (Red) and State of Charge (SOC, Green)

17:10 17:20 17:30 17:40 17:50

Time710

720

730

740

750

760

770

780

790

800

Batter y

Voltage

( V)

60

70

80

90

100

Sta

teofC

harg

eS

OC

( %)

Batt VSOC

(hh:mm)

17:10 17:20 17:30 17:40 17:50

Time3.6

3.7

3.8

3.9

4

4.1B

atter y

Cell

Volta

ge

( V)

HighestAverageLowest

(hh:mm)

During the full power discharge test at Thebarton.

The full output power of the inverter and battery tests

SoC drops 30%In 20 min

• Battery voltage is loosely correlated with state of charge.

• At the instant the inverter starts to charge the battery, the battery voltage jumps almost 10V, however the state of charge cannot instantly increase (the energy required to do so has not yet been delivered to the battery, and that energy takes time to flow into the battery).

• A large, high power battery is made from a collection of low voltage cells.

• LG CHEM battery:Each cell 3.7V14 modules make 51.8VEach rack uses 14 modules in series, 725V(Each rack uses 196 Lithium Ion cells, 91kWh energy, 126Ah @725V)

Battery power, charge and discharge limits from the battery section controller (BSC) during the full power discharge test at Thebarton.

• The instantaneous charge and discharge power limits are determined in real time by each rack’s battery management system and the battery section controller that overseas all battery racks.

• Charging of the battery starts above 100kW, but automatically drops below this as the state of charge rises.

• This drop in charge power is automatically controlled by the control system, that coordinates the inverter’s power commands in response to the battery’s state of charge.

• As state of charge rises (near 90%), the battery’s charge power limit drops. The limit doesn’t rise again until the state of charge drops.

• Because of these power limits, a grid connected battery cannot function effectively if the state of charge is too high or too low.

• Grid applications including energy arbitrage (charge when cheap, discharge when expensive), voltage support, reactive power support and peak shaving require a state of charge that allows high power charge and discharging at any time.

The instantaneous charge and discharge power limits of the battery

24/05/2019

5

Efficiency and Capacity

Inverter efficiency during the 20kW charge test.

Inverter efficiency during the full power discharge test.

5 10 15 20 25

P DC0

20

40

60

80

100

Invert

er

Effic

ienc y

(%)

(kW)

• Efficiency describes how much power and energy is lost or wasted by a system in the context of how much power and energy it delivers.

• The inverter runs below 7.4% of its rated power and demonstrates poor efficiency (high loss relative to power output) when operated at low power levels.

• The figure shows a relatively flat, high efficiency operating region from 10% rated power (>92% efficient) to 100% rated power (>97.5% efficient).

Energy Capacity

Battery voltage (Red) and State of Charge (SOC, Green).

Battery and inverter accumulated energy.

18:00 20:00 22:00 00:00 02:00 04:00 06:00 08:00 10:00 Time640

660

680

700

720

740

760

780

800

Battery

Voltage

(V)

0

10

20

30

40

50

60

70

80

90

100

Battery

Sta

teofC

harg

e(%

)

Batt VSOC

(hh:mm)

18:00 20:00 22:00 00:00 02:00 04:00 06:00 08:00 10:00

Time0

50

100

150

200

250

Energ

yO

utp

ut(k

Wh)

Battery DCInverter AC

(hh:mm)

91.5% to 0% SOC:230kWh at Inverter AC240kWh at Battery DC

• The capacity of an energy storage system is measured using an energy meter while fully discharging the battery.

• The figure shows the battery voltage and state of charge from the start of the full power discharge test (91.5% SOC), to the end of the full discharge test (0% SOC) on the following day.

• The figure shows the energy meter plots for the inverter AC connection and the battery/inverter DC connection, taken over the same time.

During the full discharge test at Thebarton

24/05/2019

6

Observation the grid-connected mode

• The plot above shows the overall grid battery operation during and after about a 40 min demonstration. Note that after the demonstration of discharge, the grid battery returns to a smart/autonomous mode where it recharges the battery then returns to an idle state

• Power was exported to reduce the load on the 11kV feeder (almost to zero power)

• The active power setpoints were manually set, but reactive power was set automatically (to control the upstream voltage at the 11kV feeder).

• The test operated very close to or at the maximum rated power of the pole mounted transformer (200kW) the grid battery was connected to.

Typical Battery Voltages and DC Power to Battery

Battery power during islanded operation of the wind tunnel fans, sampled at 5Hz (10 cycles per point). Note the harmonic power due to the bi-directional inverter operation

24/05/2019

7

• Calculated by integrating the battery power (10 minute aggregated measurements, since the integration was susceptible to small offsets).

• Negative energy when the battery has discharged, and positive energy when the battery has charged.

• The battery energy shows the depth of discharge of the battery during its autonomous operation.

The battery voltage over four months and can provide an approximateview of the batteries state of charge and depth of discharge.

The battery energy plot represents the net energy that is cycling through the battery.

Battery Voltage and Energy over 4 Months

Charging

Discharging

Preliminary Results: Battery Disconnect Event

24/05/2019

8

© 2019, all rights reserved.