Professional Documents
Culture Documents
Supercapacitor
Supercapacitor
Supercapacitor
Abstract—This paper studies the supercapacitor energy deliv- and load [5], accommodate fast and slow power components
ery capability during a constant power discharge process, which [6], and improve the power quality [7].
refers to the amount of energy delivered by a supercapacitor To exploit the supercapacitor technology, a comprehensive
when a constant power load is applied. Extensive constant power
discharge experiments are conducted using three supercapac- and in-depth understanding of its characteristics at the device
itor samples with different rated capacitances from different level is crucial. Modeling and characterization of supercapaci-
manufacturers. The relationship between the delivered energy tors have been of great interest. A variety of equivalent circuit
and the discharge power is examined. In the upper bound case models have been proposed [8]–[13]. Moreover, numerous
corresponding to a fully charged supercapacitor, the delivered frameworks have been developed to identify supercapacitor
energy increases when the discharge power decreases if the
discharge power is above a certain threshold, i.e., Peukert’s parameters and estimate supercapacitor states [14]–[17]. Re-
law applies. When the discharge power is below the threshold, cently, the impact of supercapacitor physics especially the
this law does not apply anymore. In the lower bound case charge redistribution process on various aspects of the superca-
corresponding to a partially charged supercapacitor, the delivered pacitor behavior has been investigated. Charge redistribution
energy peaks at a particular discharge power. These relationships is a relaxation process originated from the porous structure
are due to the combined effects of three aspects of supercapacitor
physics: porous electrode structure, charge redistribution, and of the supercapacitor electrodes. In addition to the physical
self-discharge. This paper also compares the bounds of the mechanisms [18] leading to charge redistribution, the effects
delivered energy and shows that the difference is significant. of this process on power management strategies in wireless
Index Terms—Supercapacitor, energy delivery capability, con- sensor networks [19], [20], supercapacitor terminal voltage
stant power load, charge redistribution, self-discharge. behavior [21]–[23], supercapacitor charge capacity [24]–[26],
and supercapacitor capacitance characterization methods [27]
I. I NTRODUCTION have been extensively studied.
Energy storage has been increasingly recognized as a crit- This paper investigates the supercapacitor energy deliv-
ical asset in many applications such as smart grid, electric ery capability during a constant power discharge process.
vehicles, wireless sensor networks, biomedical devices, and Specifically, this paper studies the upper and lower bounds
cyber-physical systems. For instance, 1676 operational or of the amount of energy delivered by a supercapacitor when
announced projects totaling a rated power of 194.02 GW have a constant power load is applied and analyzes the underly-
been reported to the DOE Global Energy Storage Database ing physical mechanisms. A constant power load is present
as of March 2018 [1]. The significant growth of global when a power converter tightly regulates its output and a
energy storage installation is due to the huge technical and destabilizing effect is introduced, which is usually referred to
economic benefits introduced by a variety of applications as the negative impedance instability [28], [29]. Therefore,
and use cases of these systems. In fact, 17 energy storage a better understanding the supercapacitor behavior during
applications grouped into five categories have been identified a constant power discharge process can facilitate designing
and analyzed [2]. Energy storage technologies are different in and implementing more effective and reliable energy storage
terms of various characteristics such as energy density, power systems. This paper adopts a methodology similar to the one
density, cycle life, leakage rate, and ramp rate [3]. In general, established in [24], [25], which is developed to study the
supercapacitors and batteries are complementary technologies: effects of supercapacitor physics on the amount of charge
high power density, low energy density, and long cycle life for delivered during a constant current discharge process. This
supercapacitors versus the opposite for batteries. Depending on paper also extends [23] in which a preliminary study of the
the application, an energy storage system may employ a single supercapacitor energy behavior is conducted.
technology or a combination of several. For example, different The remainder of this paper is organized as follows. Sec-
configurations of supercapacitor-based energy storage systems tions II and III examine the upper and lower bounds of the
[4] in which supercapacitors are employed as the sole tech- supercapacitor energy delivered during a constant power dis-
nology or combined with batteries in hybrid systems are used charge process, respectively. Section IV compares the bounds.
in microgrids to balance the power mismatch between source Section V concludes this paper.
Sample 1 2 3 30
Manufacturer Eaton AVX Maxwell
Energy (J)
29
Model HV10302R7106R SCCV60B107MRB BCAP0350
CR (F) 10 100 350 28
3 26
2.5 25 −3 −2 −1 0
10 10 10 10
Power (W)
2
(a)
Voltage (V)
300
1.5
290
1
280
0.5
Energy (J)
270
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 260
Time (s) 4
x 10
250
Fig. 1. Upper bound case: a 1 W constant power discharge experiment for
240
supercapacitor sample 2.
230 −2 −1 0 1
10 10 10 10
Power (W)
1030
by a long time constant voltage charge phase to ensure that 1020
it is fully charged. Similarly, the upper bound of the energy 1010
delivered during a constant power discharge process can be 1000
measured by discharging a fully charged supercapacitor. To 990
study the supercapacitor energy delivery capability, the three 980 −2 −1 0 1
samples listed in Table I are tested using an automated Maccor 10 10
Power (W)
10 10
1959
! % ' + $ 280
,! ,% ,' ,+ ,$
,-
"#"!$ "#&! !!#()$ %')#$ +)$"
260
*! *% *' *+ *$
&
)" !& ( + % !"""" 240
Energy (J)
220
180
I k t = Q0 , (1) 140
10
−2 −1
10
0
10 10
1
Power (W)
1960
2.8 23
2.6 22
2.4
21
2.2
Voltage (V)
Energy (J)
20
2
VT 19
1.8
V1
V2 18
1.6
V3
1.4 V4
17
V5
Energy (J)
195
1.8
190
1.6
1.4 185
1.2 180
1
0 0.5 1 1.5 2 2.5 3 3.5 4 175 −2 −1 0 1
Time (s) x 10
4 10 10 10 10
Power (W)
(b)
(b)
940
Fig. 5. Upper bound case: simulated terminal and branch capacitor voltages.
(a) Discharge power is 1.35 W. (b) Discharge power is 0.00675 W. 930
3 920
910
Energy (J)
2.5
900
2 890
Voltage (V)
1.5 880
870
1
860 −2 −1 0 1
10 10 10 10
0.5 Power (W)
(c)
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s) 4
x 10 Fig. 7. Lower bound case: relationship between delivered energy and
discharge power. (a) Sample 1. (b) Sample 2. (c) Sample 3.
Fig. 6. Lower bound case: a 1 W constant power discharge experiment for
supercapacitor sample 2.
1961
140 3
VT
V1
135 2.5
V2
V3
130 2 V4
V5
Voltage (V)
Energy (J)
125 1.5
120 1
115 0.5
110 0
−2 −1 0 1 0 10 20 30 40 50 60 70 80 90 100 110
10 10 10 10 Time (s)
Power (W)
(a)
Fig. 8. Lower bound case: simulated relationship between delivered energy 3
and discharge power. VT
V1
2.5
V2
V3
2 V4
W for sample 2, and 1.8 W for sample 3. The discharge power V5
Voltage (V)
at which the delivered energy peaks shifts to a larger value in 1.5
Sample 3
40
the simulated relationship between the delivered energy and
35
the discharge power, which is consistent with the experimental
30
results shown in Fig. 7. Specifically, the delivered energy peaks
25
at 139.38 J when the discharge power is 3 W and decreases
20
when the discharge power deviates from 3 W.
15
The effects of supercapacitor physics can be illustrated
10
using the simulation results shown in Fig. 9. After the 10
5 −3
A constant current charge process, the supercapacitor is only 10 10
−2
10
−1 0
10
1
10
Power (W)
partially charged: the five branch capacitor voltages V1 − V5
are 2.550, 1.471, 0.116, 0.001, and 0.000 V at the end of Fig. 10. Difference between upper and lower bounds of delivered energy.
this process. When a discharge power is applied, charge is
first extracted from C1 . Although V1 drops continuously, it
is still greater than V2 − V5 during the early stage of the IV. C OMPARISONS OF D ELIVERED E NERGY B OUNDS
discharge process and charge is therefore transferred from
C1 to C2 − C5 . As the discharge process continues, the This section compares the bounds of the delivered energy
charge redistribution direction is reversed during the late stage. by quantifying the difference between them as follows:
Therefore, different from the upper bound case in which Emax − Emin
charge redistribution is unidirectional (i.e., from slow branches δE = × 100%, (3)
ER
to fast branches), charge redistribution is bidirectional in the
lower bound case. When a sufficiently low discharge power where Emax is the upper bound and Emin is the lower bound.
is applied, the discharge time is significantly extended and In (3), ER is the delivered energy calculated using the rated
self-discharge contributes to a drop in the delivered energy. capacitance CR :
Therefore, the delivered energy decreases when the discharge ER = 0.5CR V12 − 0.5CR V22 , (4)
power decreases from the peaking value, which explains the
second piece of the delivered energy pattern. As for the where V1 = 2.7 V and V2 = 1.35 V. The difference between
first piece in which the delivered energy increases when the the upper and lower bounds is normalized to compare the
discharge power decreases, the mechanisms are similar to results for the three supercapacitor samples with different rated
those in the upper bound case. capacitances. The results are shown in Fig. 10.
1962
The following observations can be made. First, the differ- [9] S. Buller, E. Karden, D. Kok, and R. W. D. Doncker, “Modeling the
ence between the bounds is significant for all samples. The dynamic behavior of supercapacitors using impedance spectroscopy,”
IEEE Transactions on Industry Applications, vol. 38, no. 6, pp. 1622–
maximum difference is observed for sample 1 at 0.00675 W: 1626, 2002.
47.9% or 13.08 J while the minimum is observed for sample [10] Y. Diab, P. Venet, H. Gualous, and G. Rojat, “Self-discharge charac-
3 at 18 W: 7.6% or 72.72 J. Second, the difference depends terization and modeling of electrochemical capacitor used for power
electronics applications,” IEEE Transactions on Power Electronics,
on the discharge power. In general, a lower discharge power vol. 24, pp. 510–517, 2009.
results in a greater difference. For sample 2, the difference [11] V. Sedlakova, J. Sikula, J. Majzner, P. Sedlak, T. Kuparowitz, B. Buer-
decreases from 42.4% to 14.0% when the discharge power gler, and P. Vasina, “Supercapacitor equivalent electrical circuit model
based on charges redistribution by diffusion,” Journal of Power Sources,
increases from 0.00675 to 13.5 W. For sample 1, the difference vol. 286, pp. 58–65, 2015.
peaks at 47.9% at 0.00675 W and decreases to 11.0% at 1.35 [12] A. Szewczyk, J. Sikula, V. Sedlakova, J. Majzner, P. Sedlak, and
W. Although the difference also decreases when the discharge T. Kuparowitz, “Voltage dependence of supercapacitor capacitance,”
Metrology and Measurement Systems, vol. 23, no. 3, pp. 403–411, 2016.
power decreases, the drop is minimal: 47.6% at 0.00135 W. [13] H. Yang and Y. Zhang, “Characterization of supercapacitor models
A similar observation holds for sample 3. for analyzing supercapacitors connected to constant power elements,”
Journal of Power Sources, vol. 312, pp. 165–171, 2016.
V. C ONCLUSION [14] A. Nadeau, M. Hassanalieragh, G. Sharma, and T. Soyata, “Energy
awareness for supercapacitors using Kalman filter state-of-charge track-
This paper examines the supercapacitor energy delivery ca- ing,” Journal of Power Sources, vol. 296, pp. 383–391, 2015.
pability during a constant power discharge process. The upper [15] L. Zhang, Z. Wang, F. Sun, and D. G. Dorrell, “Online parameter
identification of ultracapacitor models using the extended Kalman filter,”
and lower bounds of the delivered energy are investigated. Energies, vol. 7, no. 5, pp. 3204–3217, 2014.
In the upper bound case, Peukert’s law applies when the [16] N. Reichbach and A. Kuperman, “Recursive-least-squares-based real-
discharge power is above a certain threshold and does not time estimation of supercapacitor parameters,” IEEE Transactions on
Energy Conversion, vol. 31, no. 2, pp. 810–812, 2016.
apply anymore if the discharge power is below the threshold. [17] H. Chaoui and H. Gualous, “Online lifetime estimation of supercapaci-
In the lower bound case, the delivered energy peaks at a tors,” IEEE Transactions on Power Electronics, vol. 32, no. 9, pp. 7199–
particular discharge power and decreases when the discharge 7206, 2017.
[18] J. W. Graydon, M. Panjehshahi, and D. W. Kirk, “Charge redistribution
power deviates from this value. The effects of three aspects of and ionic mobility in the micropores of supercapacitors,” Journal of
supercapacitor physics on the delivered energy are illustrated: Power Sources, vol. 245, pp. 822–829, 2014.
porous electrode structure, charge redistribution, and self- [19] H. Yang and Y. Zhang, “A task scheduling algorithm based on superca-
pacitor charge redistribution and energy harvesting for wireless sensor
discharge. Finally, comparisons of the bounds show that the nodes,” Journal of Energy Storage, vol. 6, pp. 186–194, 2016.
difference is significant for the three samples examined. [20] H. Yang and Y. Zhang, “Power management in supercapacitor-based
wireless sensor nodes,” in Supercapacitor Design and Applications.
ACKNOWLEDGMENT InTech, 2016, ch. 9, pp. 165–179.
[21] H. Yang, “Analysis of supercapacitor charge redistribution through
This work was supported in part by the National Institute of constant power experiments,” in Proceedings of the 2017 IEEE Power
General Medical Sciences of the National Institutes of Health & Energy Society General Meeting (PESGM 2017), 2017, pp. 1–5.
[22] H. Yang, “Bounds of supercapacitor open-circuit voltage change after
under Award 5UL1GM118979-04 and in part by California constant power experiments,” in Proceedings of the 10th Electrical
State University, Long Beach under the ORSP, RSCA, and Energy Storage Applications and Technologies (EESAT 2017), 2017, pp.
TRANSPORT programs. 1–5.
[23] H. Yang, “Impact of charge redistribution on delivered energy of
supercapacitors with constant power loads,” in Proceedings of the 2018
R EFERENCES IEEE Applied Power Electronics Conference and Exposition (APEC
[1] The DOE Global Energy Storage Database. [Online]. Available: 2018), 2018, pp. 2686–2690.
http://www.energystorageexchange.org/projects [24] H. Yang, “Estimation of supercapacitor charge capacity bounds consid-
[2] J. Eyer and G. Corey. Energy storage for the electricity grid: ering charge redistribution,” IEEE Transactions on Power Electronics,
Benefits and market potential assessment guide. [Online]. Available: vol. 33, no. 8, pp. 6980–6993, 2018.
http://www.sandia.gov/ess/publications/SAND2010-0815.pdf [25] H. Yang, “Effects of supercapacitor physics on its charge capacity,” IEEE
[3] M. Farhadi and O. Mohammed, “Energy storage technologies for Transactions on Power Electronics, p. in press, 2018.
high-power applications,” IEEE Transactions on Industry Applications, [26] H. Yang, “A study of Peukert’s law for supercapacitor discharge time
vol. 52, no. 3, pp. 1953–1961, 2016. prediction,” in Proceedings of the 2018 IEEE Power & Energy Society
[4] H. Yang, “A review of supercapacitor-based energy storage systems General Meeting (PESGM 2018), 2018, pp. 1–5.
for microgrid applications,” in Proceedings of the 2018 IEEE Power [27] H. Yang, “A revisit to supercapacitor capacitance measurement method
& Energy Society General Meeting (PESGM 2018), 2018, pp. 1–5. 1A of IEC 62391-1,” in Proceedings of the 2018 IEEE Energy Conver-
[5] H. Kanchev, D. Lu, F. Colas, V. Lazarov, and B. Francois, “Energy sion Congress and Exposition (ECCE 2018), 2018, p. in press.
management and operational planning of a microgrid with a PV-based [28] A. Kwasinski and C. N. Onwuchekwa, “Dynamic behavior and stabiliza-
active generator for smart grid applications,” IEEE Transactions on tion of DC microgrids with instantaneous constant-power loads,” IEEE
Industrial Electronics, vol. 58, no. 10, pp. 4583–4592, 2011. Transactions on Power Electronics, vol. 26, no. 3, pp. 822–834, 2011.
[6] Y. Gu, W. Li, and X. He, “Frequency-coordinating virtual impedance for [29] A. M. Rahimi and A. Emadi, “Active damping in DC/DC power
autonomous power management of DC microgrid,” IEEE Transactions electronic converters: A novel method to overcome the problems of
on Power Electronics, vol. 30, no. 4, pp. 2328–2337, 2015. constant power loads,” IEEE Transactions on Industrial Electronics,
[7] Q. Xu, J. Xiao, X. Hu, P. Wang, and M. Y. Lee, “A decentralized power vol. 56, no. 5, pp. 1428–1439, 2009.
management strategy for hybrid energy storage system with autonomous [30] D. Doerffel and S. A. Sharkh, “A critical review of using the Peuk-
bus voltage restoration and state-of-charge recovery,” IEEE Transactions ert equation for determining the remaining capacity of lead-acid and
on Industrial Electronics, vol. 64, no. 9, pp. 7098–7108, 2017. lithium-ion batteries,” Journal of Power Sources, vol. 155, no. 2, pp.
[8] L. Zubieta and R. Bonert, “Characterization of double-layer capacitors 395–400, 2006.
for power electronics applications,” IEEE Transactions on Industry
Applications, vol. 36, no. 1, pp. 199–205, 2000.
1963