Impact of the voltage fluctuation of the power supply on the efficiency of alkaline water electrolysis

Impact of the voltage fluctuation of the power supply on the efficiency of alkaline water electrolysis

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e8 Available online at www.sciencedirect.com ScienceDire...

2MB Sizes 3 Downloads 200 Views

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e8

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Impact of the voltage fluctuation of the power supply on the efficiency of alkaline water electrolysis  a *, Arp  d Bence Palota s Zsolt Dobo ros, Hungary University of Miskolc, Institute of Energy and Quality Affairs, 3515, Miskolc, Egyetemva

article info

abstract

Article history:

The efficiency of alkaline water electrolysis under DC conditions is largely dependent on

Received 5 July 2015

the voltage fluctuation of the power supply. As a result of continuously varying voltage

Received in revised form

levels the electrode reactions are under dynamic influences, affecting cell performance and

10 April 2016

gas production. This effect was investigated by conducting a series of water splitting ex-

Accepted 14 May 2016

periments where sinusoidal waves of increasing amplitude, frequency and offset was

Available online xxx

applied to a cell. Frequency ranged from 1 to 5000 Hz, signal amplitude was changed from 0 to 2 V, while the offset value (i.e. the DC component of the output signal) was varied

Keywords:

between 1.4 V and 2.8 V. A fully automated, remote-controlled measurement system was

Hydrogen production

designed, allowing for a large number of measurements. The used power supply can

Water electrolysis

generate any type of waveform signal up to 50 kHz and ±10 V with a maximum current 8 A.

Power supply

By using a fluctuating DC power source H2 flow rate and the power consumption of the

Voltage fluctuation

electrolyser can be improved, but at the same time this may lead to a drop in overall efficiency. The details of the experimental system as well as the results of the experiments are presented here. © 2016 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

Introduction Hydrogen is mostly produced from fossil resources, primarily from methane (natural gas) [1] due to cost-efficiency reasons. However, growing concerns about diminishing fossil fuel reserves and increasing gas prices drive attention to hydrogen production. The most common arguments against water decomposition using electrical power are of economic nature. Net efficiency of 56e73% is common with current electrolytic hydrogen production methods [1,2], however the overall efficiency is much lower if the losses of electricity generation

from primary or renewable fuels are also taken into consideration. In this case the overall efficiency rate would hardly reach 25e40% [3]. The electrolytic process can be enhanced by manipulating external factors, i.e. by subjecting the cell to a super gravity field [4], to ultrasonic waves [5] or to permanent magnets [6]. If, on the other hand, a fluctuating DC source is applied instead of steady DC voltage, efficiency decline might occur in some cases. The conventional DC electrolysis of water involves generation of hydrogen gas at the cathode and oxygen gas at the anode. The electrode reactions are typically described as follows [7]:

* Corresponding author.   ), [email protected] (A.B.  s). E-mail addresses: [email protected] (Z. Dobo Palota http://dx.doi.org/10.1016/j.ijhydene.2016.05.141 0360-3199/© 2016 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.  Impact of the voltage fluctuation of the power supply on the efficiency of alkaline  Z, Palota  s AB, Please cite this article in press as: Dobo water electrolysis, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.05.141

2

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e8

cathode: 2H2O þ 2e⁻ / H2 þ 2OH⁻ 2H⁺ þ 2e⁻ / H2 anode: 2OH⁻ / 0.5O2 þ H2O þ 2e⁻ H2O / 0.5O2 þ 2e⁻ þ 2H⁺ Gas formation can be induced by applying a potential difference between the cell electrodes. In theory, the thermodynamic decomposition voltage of water at 298 K and 1 atm is 1.23 V, however, due to reaction overpotentials and resistance losses (ohmic voltage drop), higher actual cell voltages should be applied [8]. Gas starts to form at a voltage level of 1.65e1.7 V and most industrial cells are operated from 1.8 to 2.6 V [8,9]. In order to create a potential difference between the electrodes AC/DC or DC/DC power supplies are typically used. One of the main properties of power supplies is the voltage ripple (noise), meaning that the voltage level as a function of time is not constant. Reduced or practically eliminated voltage ripple is an important requirement, although it is difficult to achieve with increased power. In water electrolysis any deviation from perfectly uniform DC voltage might affect the electric power consumption, the intensity of gas production and the efficiency of water splitting. This means, that using a power supply with a significant ripple can decrease the efficiency of the water electrolysis. Inversely, by smoothing the voltage fluctuation of the DC power supply the electrolysis efficiency can significantly increase. Although several papers were published on applying impulse voltage or interrupted direct current to the cell [10e13], according to some authors the available information about the effects of the applied voltage waveforms are relatively low [14,15]. In 2009 a group of scientists took an alternative approach, studying the influence of different types of power supply on the efficiency of alkaline water electrolysis [16]. It was shown, that the efficiency of water electrolysis can be changed by using different power supply topologies with different output voltage and current waveforms. The ripple of the available power supplies from 5 kW can reach 5e7% [17], in some cases it can even go up to 10% [18]. The main goal of this paper is to generate sinusoidal waveforms with various frequency, amplitude and offset values, and apply them to the electrolytic cell, modelling the ripple of power supplies. During the experiments cell voltage, cell current and the gas production were measured. Since the number of parametric combinations was obviously far too high to execute the experiments manually (within a realistic time scale), a special experimental system was set up, enabling fully automated control, measurement and data logging.

Fig. 1 e Schematic illustration of the measurement system.

Table 1 e Basic output parameters of the power supply. Output parameter Voltage, V Electric current, A Frequency, kHz

Minimum value

Maximum value

10 0 0

þ10 8 50

device type USB-6259. The NI device generates a signal programmed in LabVIEW software (National Instruments, 11.0 version, 2011). This system can generate any type of waveforms between the criteria showed in Table 1. The power supply can be operated either in galvanostatic or potentiostatic mode, i.e., it is able to generate voltage or current waveforms, according to the operational amplifier topology [19]. The power supply is connected to the electrolytic cell shown in Fig. 2. The cell is basically a gas-tight vessel with two electrodes attached to the lid. Next to the electrodes are two gas outlets, a T-type Thermocouple (used as a sensor to measure electrolyte temperature) and the outlet pipes of the cooling cycle (mounted on the lid). Two stainless steel electrodes (dimensions: 1.5  25  25 mm; material: EN 1.4307)

Measurement system setup A schematic illustration of the measurement system is seen in Fig. 1. The central unit is the power supply, containing a high current operational amplifier OPA549 (manufactured by Texas Instruments) [19] controlled by National Instrument (NI)

Fig. 2 e Schematic illustration of the electrolytic cell, a) cathode, b) anode, c) to pressure sensor, d) to valve, e) cooling water inlet, f) cooling water outlet.

 Impact of the voltage fluctuation of the power supply on the efficiency of alkaline  Z, Palota  s AB, Please cite this article in press as: Dobo water electrolysis, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.05.141

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e8

were placed in the cell, their distance being 40 mm. 500 ml of 30 wt.% potassium hydroxide aqueous solution was used as electrolyte. The gas bubbles form at the electrode surfaces cause pressure increase in the cell. The rate of pressure increase is directly proportional to the flow rate of the gas produced. Pressure was measured using a pressure sensor type MPX5010DP (manufactured by Freescale Semiconductor), capable of measuring differential pressure up to 10 kPa at a sensitivity of 450 mV/kPa [20]. The rate of pressure increase was calibrated with an independent flow meter. While the pressure sensor was fixed at one of the gas outlets, the second gas outlet was connected to an electromagnetic valve. When the valve is open the pressure in the cell decreases to the ambient pressure and H2 þ O2 gas are released into the environment. The pressure increase in one experiment is negligible (the volume of the produced gas is very small relative to the available space) and consequently it does not have any effect on the reaction outcome. However, considering a long series of experiments, to prevent pressure build-up the formed gas should be released (through the valve) after each measurement in order to avoid distortion of the measurement results. The position of the valve (open/closed) can be controlled at the digital output terminal of the signal generator module. The fully automated measurement system allows for a large number of water splitting experiments. The system is also remote-controlled, allowing for changes in measurement series or obtaining information about the results and actual cell conditions through the internet.

Measurement sequences A set of multiple measurement series were designed, coupling sinusoidal signal waveforms (Fig. 3) of varying frequency (f) and amplitude (a) at 8 different DC offset levels (UDC). The offset value was increased stepwise from 1.4 to 2.8 V (in 0.2 V steps). Signal amplitude went up from 0 V to 2 V in 0.2 V steps. Signal frequency was raised stepwise by 4 Hz from 1 to 97 Hz, then by 100 Hz from 200 to 5000 Hz. 814 separate measurements were taken by the automated measurement system for each (preset) offset value, as illustrated in the flow chart below (Fig. 4). Each signal form was

3

Fig. 4 e Flow diagram of the automated measurement process.

delivered to the cell for 15 s while the following data were measured and recorded by the software: - cell voltage, - cell current and - cell pressure (sampling rate was 1000 sample/sec up to 97 Hz; then 50,000 sample/sec from 200 Hz). Five seconds after each measurement the magnetic valve was opened for 10 s to depressurize the cell and restore equilibrium pressure in the gas space. Then the valve was closed and after 30sec pause the next measurement was taken. Thus a total of 6512 measurements were performed over a period of almost five days. About 90 GB of data were collected during the 6512 measurements. Data analysis was performed by a custom software (developed in C#). The two most important cell parameters were calculated during the analysis:  the rate of pressure rise, Pa/sec;  average power consumption, W.

Fig. 3 e Characterization of the sine waveform.

In each case, pressure evolved linearly inside the cell. First, using linear regression, then the slope of the pressure line was computed and fitted to the data points to determine the rate of pressure rise. The coefficient of determination typically fell between 0.98 and 0.99.

 Impact of the voltage fluctuation of the power supply on the efficiency of alkaline  Z, Palota  s AB, Please cite this article in press as: Dobo water electrolysis, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.05.141

4

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e8

The electrical performance of the cell (P) was computed using the following formula: Pn P¼

i¼1 jUn $In j

(1)

n

where:  n e the number of datapoints;  Un e actual cell voltage, V;  In e actual cell current, A. Water loss was negligible and practically had no effect on electrolyte composition or volume, therefore this parameter was considered constant. All measurements were taken at a temperature of 25.5 ± 1  C at ambient pressure.

Results Fig. 5 displays the production rates (gas flowrates) obtained by applying periodically changing sinus-shaped voltage signals to the cell electrodes at different parametric setups. For the gas flows measured at 8 different DC offset values, similar trends can be observed: the produced flows tend to increase with increasing amplitude and decreasing frequency. Given that a ¼ 0 V represents the baseline case (i. e. steady DC), it is obvious from Fig. 5 that the coupling of sinus-shaped signals to any given DC level yields positive gas flow changes (relative to the offset value). Voltage waveforms applied to the cell cause fluctuation also in electric current. Given the fact that the hydrogen produced in the electrolyser is directly proportional to the mean current supplied, Fig. 5 clearly shows how the electric current

is changing by varying the voltage setups. Further investigations are needed in order to observe the connection between the voltage and current fluctuation. Upon increasing the frequency of sinusoidal periodic waveforms, a limit will be approached, which practically denotes to the flow rate of gas obtained under steady DC output equivalent to the offset value of the periodic signal. Though this cut-off value is reached in any case, the larger the amplitude of the signal, the higher frequencies should be applied to obtain DC-equivalent flowrates. By increasing the amplitude while keeping signal offset at 1.8 V and frequency at 1 Hz, an immediate improvement in volumetric performance can be achieved. Any change to the offset value (whether positive or negative) implies higher amplitudes to maintain gas production intensity. The amount of electricity required for water decomposition is represented against the parameters of sinusoidally varying voltage in Fig. 6. The same trends can be observed for offset change as in the previous case (see Fig. 5): with higher offset values, larger signal amplitudes should be applied to attain improved cell response. While the gas production rate is directly proportional to the electric current, the power consumption calculations involves also the cell voltages (see Eq. (1)). As a result of this in the power vs. frequency and amplitude graphs the voltage amplitudes plays an important role. The gas production at a given offset value does not change in many cases, however, the power consumption at a given offset value is affected mainly by the voltage amplitude, and the waveform frequency has only minor impact. It means there are a lot of setups where the efficiency loss of the electrolyser is higher compared to the true DC electrolysis.

Fig. 5 e Flow rates of the produced gas as a function of amplitude and offset values.  Impact of the voltage fluctuation of the power supply on the efficiency of alkaline  Z, Palota  s AB, Please cite this article in press as: Dobo water electrolysis, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.05.141

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e8

5

Fig. 6 e Power input/consumption as a function of frequency, amplitude and DC signal offset.

Efficiencies were calculated from cell performance and gas flow rates, as shown in Fig. 7. Each data point corresponds to a given cell performance, composed of a theoretical (Videal) and a measured (Vmeasured) value of flow rate. Efficiency is calculated as the ratio of the two [3]: h¼

100 Vmeasured Videal

(2)

where Videal was calculated using Faraday's law: Videal ¼

  Vm P 1 1 þ U0 F zH2 zO2

(3)

where Vm is the molar volume of ideal gases at normal conditions (22,410 cm3/mol), P is the measured power consumption of the cell (see eq. (1)), U0 is the theoretical thermodynamic decomposition voltage of water (1.23 V), F is Faraday's number (96,485 C/mol) and z is charge number (2 for H2 and 4 for O2). All flow rates were evaluated at normal state (0  C, atmospheric pressure). If frequency is kept low, efficiency will increase with signal amplitude up to an offset value UDC ¼ 1.8 V. At offset values higher than UDC ¼ 2 V, cell performance tends to decrease with increasing amplitude. When UDC is 2 V, efficiency is tightly frequency-dependent, frequency increase leads to efficiency decline even with small amplitudes (0.1e0.2 V). Higher offset values allow for a wider range of amplitudes where cell efficiency is left unaffected by frequency change.

The highest productivity of the presented electrolyser was 5.1 kW h/Nm3 H2, which was measured at 2.2 V DC. Efficiency degradation as a percent rate of baseline (DC) cell efficiency is displayed in Fig. 8. Four distinct levels of efficiency decrease are plotted against the offset and amplitude of the signal at a given frequency. Note that the initial offset value in the graph is 2 V, since practically no cell efficiency occurs up to 1.8 V DC. If frequency is kept at 5 Hz and offset is greater than 2 V, then a minimum of 0.25 V signal amplitude should be used in order to make 2% efficiency loss. However, only 0.1 V amplitude can cause 20% efficiency loss at 5 kHz and 2 V offset. For the power supply side, it is possible to predict from the Figure what cell efficiency can be enhanced by using steady DC power supply instead of periodically changing voltage forms. Consider, for example, a power supply output with sinusoidal waveforms of 49 Hz frequency, 0.1e0.15 V amplitude and 2 V DC offset value. In this case, up to 5% efficiency improvement can be reasonably expected from the results (upon proper smoothing of the voltage supply).

Conclusion The voltage fluctuation of the applied power source does not necessarily cause significant drop in the efficiency of water decomposition, compared to the baseline efficiency obtained

 Impact of the voltage fluctuation of the power supply on the efficiency of alkaline  Z, Palota  s AB, Please cite this article in press as: Dobo water electrolysis, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.05.141

6

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e8

Fig. 7 e Cell efficiency as a function of signal parameters (frequency, amplitude, and DC offset).

under steady DC. Applying a DC voltage level of e.g. 2.8 V to the cell, efficiency will not decline by more than 2% under sinusoidal voltage fluctuation up to 0.4 V amplitude over a frequency range of 1 Hze5000 kHz. The same holds for fluctuation with an amplitude of 0.2 V under 2.4 V applied offset. However, when the power supply output is set at 2 V (DC offset), even 0.1 V signal amplitude is to be precluded in order to avoid efficiency loss. We have shown that lower frequencies and higher amplitudes enhance gas production yield (gas flow rate). Applied signal frequency can be increased up to a cut-off frequency value, beyond which the flow rate of the produced gas will be the same as the yield produced by splitting water with normal DC voltage equivalent to the signal DC offset. By using a fluctuating DC power source, gas flow rate and cell

performance can be improved, nonetheless this may lead to a drop in overall cell efficiency. As the results show, the efficiency of the electrolytic cell can be enhanced by simple voltage smoothing if the power supply has a significant ripple. Enhancement rate will depend on cell configuration and the parameters of the output signal (frequency, amplitude and DC offset value). The excessive amount of measurement data required the design and assembly of a special experimental and measurement system. By the use of this system fully automated execution of multitudinous water splitting experiments were carried out within a short period of time. Similar measurement sequences using triangular, square and sawtooth signal waveforms are planned for the near future.

 Impact of the voltage fluctuation of the power supply on the efficiency of alkaline  Z, Palota  s AB, Please cite this article in press as: Dobo water electrolysis, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.05.141

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e8

7

Fig. 8 e Amplitude values for percentage rates of cell efficiency decrease as a function of offset and frequency parameters.

Acknowledgement  This research was realized in the frames of TAMOP 4.2.4. A/211-1-2012-0001 “National Excellence Program e Elaborating and operating an inland student and researcher personal support system convergence program” The project was supported by the European Union and co-financed by the European Social Fund. This research was carried out in the framework of the Center of Excellence of Sustainable Resource Management at the University of Miskolc.

references

[1] Holladay JD, Hu J, King DL, Wang Y. An overview of hydrogen production technologies. Catal Today 2009;139:244e60. http://dx.doi.org/10.1016/j.cattod.2008.08.039. [2] Turner John, Sverdrup George, Mann Margaret K, Maness Pin-Ching, Kroposki Ben, Ghirardi Maria, et al. Renewable hydrogen production. Int J Energy Res 2008;32(5):379e407. http://dx.doi.org/10.1002/er.1372. [3] The electrolysis of water (2015.02.25.) http://www.theionspa. com/page/569/electrolysis.

[4] Wang Mingyong, Wang Zhi, Guo Zhancheng. Water electrolysis enhanced by super gravity field for hydrogen production. Int J Hydrogen Energy 2010;35:3198e205. http://dx.doi.org/10.1016/j.ijhydene.2010.01.128. [5] Li Sheng-De, Wang Cheng-Chien, Chen Chuh-Yung. Water electrolysis in the presence of an ultrasonic field. Electrochimica Acta 2009;54:3877e83. http://dx.doi.org/ 10.1016/j.electacta.2009.01.087. [6] Lin Ming-Yuan, Hourng Lih-Wu, Kuo Chan-Wei. The effect of magnetic force on hydrogen production efficiency in water electrolysis. Int J Hydrogen Energy 2012;37:1311e20. http://dx.doi.org/10.1016/j.ijhydene.2011.10.024. [7] Marini Stefania, Salvi Paolo, Nelli Paolo, Pesenti Rachele, Villa Marco, Berrettoni Mario, et al. Advanced alkaline water electrolysis. Electrochimica Acta 2012;82:384e91. http://dx.doi.org/10.1016/j.electacta.2012.05.011. [8] Nikolic Vladimir M, Tasic Gvozden S, Maksic Aleksandar D, Saponjic Djordje P, Miulovic Snezana M, Kaninski Milica P Marceta. Raising efficiency of hydrogen generation from alkaline water electrolysis e energy saving. Int J Hydrogen Energy 2010;35:12369e73. http://dx.doi.org/10.1016/ j.ijhydene.2010.08.069. [9] Wang Mingyong, Wang Zhi, Gong Xuzhong, Guo Zhancheng. The intensification technologies to water electrolysis e a review. Renew Sustain Energy Rev 2014;29:573e88. http://dx.doi.org/10.1016/j.rser.2013.08.090. [10] Mazloomi Kaveh, Sulaiman Nasri, Ahmad Siti Anom, Yunus Nurul Amziah Md. Analysis of the frequency response

 Impact of the voltage fluctuation of the power supply on the efficiency of alkaline  Z, Palota  s AB, Please cite this article in press as: Dobo water electrolysis, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.05.141

8

[11]

[12]

[13]

[14]

[15]

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e8

of a water electrolysis cell. Int J Electrochem Sci 2013;8:3731e9. Shimizu Naohiro, Hotta Souzaburo, Sekiya Takayuki, Oda Osamu. A novel method of hydrogen generation by water electrolysis using an ultra-short-pulse power supply. J Appl Electrochem 2006;36:419e23. http://dx.doi.org/10.1007/ s10800-005-9090-y. Vanags Martins, Kleperis Janis, Bajars Gunars. Water electrolysis with inductive voltage pulses. Electrolysis, Chapter 2. InTech; 2012, ISBN 978-953-51-0793-4. http://dx.doi.org/ 10.5772/52453. Shaaban Dr Aly H. Pulsed DC and anode depolarization in water electrolysis for hydrogen generation. HQ air force civil engineering support agency; 1994. final report. Mazloomi SK, Sulaiman Nasri. Influencing factors of water electrolysis electrical efficiency. Renew Sustain Energy Rev 2012;16:4257e63. http://dx.doi.org/10.1016/j.rser.2012.03.052. Mazloomi Kaveh, Sulaiman Nasri b, Moayedi Hossein. Electrical efficiency of electrolytic hydrogen production. Int J Electrochem Sci 2012;7:3314e26.

[16] Ursu´a Alfredo, Marroyo Luis, Gubı´a Eugenio, Gandı´a Luis guez Pedro M, Sanchis Pablo. Influence of the M, Die power supply on the energy efficiency of an alkaline water electrolyser. Int J Hydrogen Energy 2009;34:3221e33. http://dx.doi.org/10.1016/j.ijhydene.2009.02.017. [17] Thyristor technology: http://www.americanplatingpower. com/pdfs/americanplating-mta-series.pdf (10.05.2015). [18] Contreras A, Guiradoa R, Veziroglu TN. Design and simulation of the power control system of a plant for the generation of hydrogen via electrolysis, using photovoltaic solar energy. Int J Hydrogen Energy 2007;32:4635e40. http://dx.doi.org/10.1016/j.ijhydene.2007.07.006. [19] OPA549 type operational amplifier, technical datasheet: http://www.ti.com/lit/ds/symlink/opa549.pdf (02.02.2015). [20] MPX5010DP type pressure sensor technical datasheet: http://www.freescale.com/files/sensors/doc/data_sheet/ MPX5010.pdf (16.02.2015).

 Impact of the voltage fluctuation of the power supply on the efficiency of alkaline  Z, Palota  s AB, Please cite this article in press as: Dobo water electrolysis, International Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.05.141