**8. Bench-test set-up and procedure**

Two packs of RLIB are shown in **Figure 3**. Both of them are originally applied in pure EVs. After running on board for several years, they are used as experimental targets in this study, assembled with EMS, and installed in a test bench to simulate RLIB at WWTP.

**9. Verification of RLIB pack with EMS**

accurately predicts the response of RLIB.

In the bench test, the first case of LiMnNiCoO<sup>2</sup>

and single LiFePO<sup>4</sup>

Molicel Module 10.96 V

**Table 3.** Specification of RLIB.

EME335-I403 (18650AG, 3S35P)

To consider a real severe case, the current draw of the pattern of electricity is imposed on the RLIB pack [41, 42]. As shown in **Figure 13**, the accuracy of simulation with RLIB analytical module is examined by comparing with the measured results. The simulation with assumed linear VOC yields the deviation from the measured voltage curve. Otherwise, the simulation

Pishuang Cell 38.4 Ah 3.2 V 400013201 21 cells 60 65.75 84.23

**(Wh/kg)**

3 modules 100 3.27 × 3

Simulation results regarding voltage drop of a single RLIB pack in 100(s) under random load current is compared with the other case of RLIB pack connected with UC and active controlled by EMS (**Figures 14** and **15**). Effect of active controlled by EMS represented in DoD is

of DoD with/without EMS under constant c-rate discharging. RLIB in active control of duty cycle 60% (solid line) shows the more stable and limit DoD than a single RLIB pack (dash line). Through real-time simulation by monitoring DoD, we optimize the best control duty of 60%. Here, IR of the RLIB pack plays an essential role in the distribution of DoD. To examine the control strategy even further, LiFePO4 RLIB is utilized as the DoD results (**Figure 19**). The effectiveness of EMS (solid line) is realized in comparison with the cases without EMS (point line)

RLIB (dash line). To consider the stable DoD distribution of RLIB by using

\*IR at both cases is shown

RLIB pack in **Figure 18** shows the comparison

**IR(mΩ)**

Reused Lithium-Ion Battery Applied in Water Treatment Plants

**Pack Total\***

http://dx.doi.org/10.5772/intechopen.76303

173

not obvious. However, the energy consumption estimated from I<sup>2</sup>

**Figure 13.** Comparison of simulation and measured results (upper: current; down: voltage) [41].

in **Figures 16** and **17**, and EMS decreases 26% heat loss of RLIB.

**Item Unit Energy density** 

\*Total IR is composed of internal resistance + harness resistance + fixture resistance.

Two types of LIB cells with a large difference in IR are employed in this study, and the specifications are listed in **Table 3**. An automated test bench with rated voltage and current of 500 V/450 A is utilized for the test. The initial rated voltage of RLIB is 70 V. A power pattern converted from the daily usage of electricity in WWTP is programmed into the machine for discharge/charge operation. In this study, all components are integrated in the laboratory, and the pattern of electricity is chosen for simulating the intermittent charging/discharging cycle of renewable energy and power accumulation due to the lack of in-situ energy consumption data. The duty cycle, current, and voltage of the RLIB terminal are monitored by the EMS. A total of 21 cells of LiFePO4 RLIB and three modules of LiMnNiCoO2 RLIB are modularized into two individual packs. A test case of RLIB connected with EMS is shown in **Figure 12**.

**Figure 12.** Implementation of RLIB with EMS and auxiliary physical battery (left: EMU; Central: LiFePO<sup>4</sup> LIB pack; and right: auxiliary physical battery).


**Table 3.** Specification of RLIB.

excitation (PE). It enhances the application of this method for power systems. **Figure 11** shows one example for examining the algorithm by using adaptive control observer to estimate VOC and IR through the adaptive control approach. Estimation of SOH-sensitized IR can converge

Two packs of RLIB are shown in **Figure 3**. Both of them are originally applied in pure EVs. After running on board for several years, they are used as experimental targets in this study,

Two types of LIB cells with a large difference in IR are employed in this study, and the specifications are listed in **Table 3**. An automated test bench with rated voltage and current of 500 V/450 A is utilized for the test. The initial rated voltage of RLIB is 70 V. A power pattern converted from the daily usage of electricity in WWTP is programmed into the machine for discharge/charge operation. In this study, all components are integrated in the laboratory, and the pattern of electricity is chosen for simulating the intermittent charging/discharging cycle of renewable energy and power accumulation due to the lack of in-situ energy consumption data. The duty cycle, current, and voltage of the RLIB terminal are monitored by the EMS. A total of 21 cells of LiFePO4 RLIB and three modules of LiMnNiCoO2 RLIB are modularized into two individual packs. A test case of RLIB connected with EMS is shown in **Figure 12**.

assembled with EMS, and installed in a test bench to simulate RLIB at WWTP.

**Figure 12.** Implementation of RLIB with EMS and auxiliary physical battery (left: EMU; Central: LiFePO<sup>4</sup>

right: auxiliary physical battery).

LIB pack; and

into a stable measured value in about 600(s).

172 Energy Systems and Environment

**8. Bench-test set-up and procedure**
