Open Collections

UBC Undergraduate Research

Final recommendation report for oxyhydrogen generator analysis and optimization Sobie, Cameron Jan 10, 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52966-Sobie_Cameron_APSC_479_2011.pdf [ 2.78MB ]
JSON: 52966-1.0074446.json
JSON-LD: 52966-1.0074446-ld.json
RDF/XML (Pretty): 52966-1.0074446-rdf.xml
RDF/JSON: 52966-1.0074446-rdf.json
Turtle: 52966-1.0074446-turtle.txt
N-Triples: 52966-1.0074446-rdf-ntriples.txt
Original Record: 52966-1.0074446-source.json
Full Text

Full Text

Final Recommendation Report for Oxyhydrogen Generator Analysis and Optimization Cameron Sobie∗ Faculty of Applied Science, University of British Columbia, Vancouver, B.C.,V6T 1Z4 , Canada Project Sponsor: Bernhard Zender Team Number 1054 (Dated: January 10, 2011)  Abstract Oxyhydrogen gas produced via electrolysis was performed for the first time immediately following the invention of electrolysis. It is used on the industrial scale for producing high purity hydrogen, and has become the subject of increasing scientific interest with the possiblity of commercially feasible hydrogen fuel cells. Oxyhydrogen gas also remains an effective high temperature flame source for hobbyist and specialty applications. Initially, this project aimed to build and optimize an electrolytic cell with the aims of producing useful amounts of oxyhydrogen gas. After construction of the system, it became obvious that the high power requirements to produce useful volumes of gas made this goal unfeasible. As a result, the project aim shifted to a scientific investigation of the effect of pulse width modulated power on electrolysis with the goal of improving efficiency. After numerous measurements and attempts at optimization, the experimental setup was unable to produce sufficiently accurate results to make any claims. Numerous recommendations are made should a group choose to improve the experiment setup and retry the experiment.  ∗  i  Contents  I. Introduction  1  A. Background and Motivation  1  B. Project Objectives  4  C. Organization  5  II. Theory  6  A. Safety  6  B. Gas Production Volume  6  C. Electrochemisty  6  D. Physical Effects  7  III. Experimentation  9  A. Experimental Setup and Methods  9  B. Low Power Analysis  10  C. Pulse Width Modulation  12  D. Experimental Error  14  IV. Project Deliverables  16  V. Conclusion  17  VI. Recommendations  18  A. Cell Design  18  B. Support Apparatus  18  C. Electrochemistry  19  D. Sensor Calibration  19  E. Power Supply  20  F. Pulse Width Modulation  20  VII. APPENDIX A  21  References  22  ii  List of Figures 1. Effects of cathodic corrosion.  2  2. Corroded flow adaptor.  3  3. CAD model of electrolytic cell.  4  4. Experiment apparatus setup.  10  5. Low power behaviour.  11  6. Closeup of low power behaviour.  12  7. Pulse width modulation results.  13  8. Fluctuations in gas sensor output.  14  9. Honeywell senor calibration curve.  15  iii  I.  INTRODUCTION A.  Background and Motivation  Electrolysis of water was first performed by Nicholson and Carlisle after the invention of voltaic piles in 1800 [3] and gained further popularity with the newer sources of electricity. Electrolysis remains a standard method of producing high purity hydrogen, though it is much more cost feasible to produce large volumes of hydrogen from fossil fuel hydrocarbons. Little research has occurred over the past several decades on strictly the efficiency of electrolysis of water in industrial practice; the process simply requires platinum electrodes and enormous amounts of power to split water into hydrogen and oxygen gas at a reported efficiency from 40 − 60%. However, this process is also gaining more interest in the scientific community with the possibility of hydrogen fuel cells becoming viable sources of energy. This new research thrust is largely on very specific, rare, and specialized materials [9], and photoassisted electrolysis, a process involving electrolysis powered by the solar power itself, not electricity[12]. Finding any methods to improve the efficiency of generating oxyhydrogen gas by the electrolysis of water could be enormously valuable in the future if hydrogen fuel cells become sufficiently developed for common commercial use. Oxyhydrogen gas (2H2 + O2 ) has long been used for welding [10] and other high temperature applications such a melting quartz and synthesizing gemstones [8]. Before the use of oxyacetylene welding, oxyhydrogen provided the highest temperature flame (2800C ◦ ) [2] and was the only method of working refractory metals at the time of its invention. Though it has fallen out of use in industrial welding because of superior alternatives such as MIG, TIG, and oxyacetylene welding, oxyhydrogen gas continues to be of use in synthesizing the stone corundum, a gem used in watch-making, with the Verneuil Process [2] and in hobbyist applications. Numerous unverified claims of automotive engine efficiency improvement due to oxyhydrogen gas addition also exist. Difficulty arises in producing oxygen and hydrogen gas with electrolysis because of the reduction-oxidation potentials in the reaction. Two half reactions occur: And at the cathode, 2H2 O(l) → O2 (g) + 4H + (aq) + 4e−  1  (1)  At the anode, 4H + (aq) + 4e− → 2H2 (g)  (2)  The reduction potential is Ered = 0.00V which does not cause issue, but the oxidation potential at the anode is Eox = −1.23V . This oxidation potential is greater than what is required to reduce most metals, including the constituent metals of the stainless steel plates used in the project. As a result, instead of only producing oxygen gas, metal ions are pulled out of the electrodes and into the solution. Below, Figure 1 shows the effect on stainless steel of cathodic erosion.  FIG. 1: Effects of cathodic erosion, with the cathode on the left and anode right. Unaffected stainless steel can be seen along the perimeters of the plates.  This electrochemical action is even more obvious on a brass flow modifer that was used in the cell; though the flow adapter was not electrified, it was placed in an electric field from the electrode plates, which caused electrolysis to occur on the surface as in Figure 2. In fact, there are only a select few metals that will not suffer cathodic erosion. Two 2  FIG. 2: Corroded brass flow adaptor. The anode is end is in the foreground.  that are guaranteed to work are gold and platinum, and other specialty alloys may function under specific pH conditions. For this project, gold foil and gold plating were investigated as possible methods to avoid corrosion of the cell electrode plates. Gold foil could not be obtained in sufficiently large sheets to coat a single electrolytic plate. This makes it difficult to guarantee a waterproof seal around the plate. Gold plating creates a seamless deposit of gold on the electrolytic plates; however, local plating shops quoted $60 per plate, which was beyond the budget of this project when using nine plates. As a result, raw stainless steel plates were used in the cell for this project. Electrolytic cells are very simple devices. Strictly speaking, the requirements are a conductive anode and cathode, which are immersed in a conductive solution. Figure 3 shows a CAD drawing of the cell used in this project. Any further modifications are implemented to improve usability and efficiency. The cell for this project uses stainless steel plates as its anode and cathode, mounted within a Lexan enclosure. Using a sodium hydroxide electrolyte is ideal because it is highly conductive, does not interfere with the reaction or leave deposits on the plates. However, the caustic effects of sodium hydroxide posed a threat to the components of the electrolytic cell support system as well as to students in the Project Lab, so a sodium bicarbonate electrolyte was selected instead. The cell itself has an opening at the bottom of the assembly allows water to flow in, and a top opening allows water and oxyhydrogen gas to flow out. The gas/water mixture flows into a separator tank shown in  3  Figure 4, which separates the water and gas.  FIG. 3: Solidworks CAD of electrolytic cell by Bernhard Zender.  B.  Project Objectives  Initially, this project was to investigate possible optimizations to the electrolytic cell configuration and to construct an oxyhydrogen production system. A gas/liquid separation vessel was made from Lexan and a PTFE water tank was purchased. All components were fitted with brass hose connections and clamped to ensure a robust seal for all plumbing in the system. Once a solid system was established, the effectiveness and economic feasibility in coating the electrode plates of the cell with gold was investigated. After all construction was completed and rudimentary testing began with a high current DC power supply, it became readily apparent that to produce amounts of oxyhydrogen gas that could be used for practical purposes, the cell demanded an enormous amount of power. At the maximum power possible from the equipment in the Project Lab, one kilowatt produced under three liters of gas per minute, which is not enough for any useful purpose. In light of this challenge, the scope of the project changed dramatically. The power 4  requirements of a oxyhydrogen production unit would incur very high costs, so the project shifted to a scientific analysis of the effect of a pulse width modulated power source on the production efficiency of the gas. The duty cycle and modulation frequency were to be varied while the standard cubic centimeters of gas per watt is measured.  C.  Organization  This report will describe all aspects of this project. The theory concerning the safety of generating oxyhydrogen gas, the of conductance of an ionic two phase solution, the anticipated gas generation for a given current, and electrochemical effects is presented. Following this, the experiment setup is described, should another group choose to repeat the experiment. The results of the testing are presented after this, followed by the Conclusion and Recommendations.  5  II.  THEORY A.  Safety  This device and associated experimentation was performed in the Project Lab; given that pure oxyhydrogen gas is highly explosive when in stoichiometric proportions, it warrants analysis to ensure lab safety during the experiment. The dangerous component of the gas released is hydrogen, as oxygen is in sufficient quantities to allow combustion in open air. The ignition temperature of gaseous hydrogen is in the range of 800 − 1000[K] [7] with a minimum required concentration of 22 to 26% [13]. Furthermore, according to Graham’s Law of gas transport, the diffusion rate of a gas is inversely proportional to the square root of its molecular mass. This means that hydrogen is the fastest diffusing gas molecule. In summary, it is safe to operate and to vent this device in the Project Lab at the gas production levels of the experiment, so long as direct flame is not introduced to the gas vent or a within reasonable proximity.  B.  Gas Production Volume  Given the equations of dissociation in the Background, one can predict the volumes of gas that will be produced for a certain current and time. One mol of gas occupies 24.5[L] −  at 25 ◦ C and 101.3[kP a]. Therefore At 1[A] of current or 6.24151 × 1018 [ es ], the electrodes will produce 4.68 × 1018 molecules of gas per second, or 1.896[ mL ]. s This process demands much current to produce usable amounts of gas. While it is inevitable that to produce more gas one will need more power, it is possible to raise the voltage requirement instead of current. Inserting additional plates will produce an according increase in gas production while increasing the voltage requirements, but maintaining the current requirement. This is because the same current passes through all plates, and simply requires more ”force” to do so.  C.  Electrochemisty  The amount of voltage required for the reaction to occur is found in an oxidation-reduction potential table. However, the values listed are purely theoretical. In reality, a reaction 6  overpotenial must be added to the listed oxidation/reduction potential, which accounts for an increase in the required potential due to interactions between the electrode material and the gas produced at the electrode. This value is 0.90[V ] for oxygen and hydrogen on iron electrodes, which is the main constituent metal of the stainless steel plates of the cell. In commercial applications, electrolysis of water is performed with platinum electrodes resulting in the minimum possible reaction overpotential of 0.84[V ].  D.  Physical Effects  Significant theory exists in the area of the conductivity of solutions as temperature, species concentration, and bubble concentration varies; however, in the area specifically regarding electrolysis, little exists that could help in predicting any effects of PWM power. At the electrode plate, many things are happening simultaneously: as water is turned into gas, bubbles form on the plate, decreasing the immersed area of the plate and increasing the effective ohmic resistance of the solution as well as causing a bubble overpotential effect [11]. As this happens, the concentration of conductors in the solution also drops, which causes concentration overpotential effects[4]. On top of these micro-scale effects, the flow rate of the solution through the cell has a significant effect as it directly affects the bubble density within the cell as well as local electrolyte concentration. Using the research on two phase solutions and data on the conductivity of solutions, it is possible to find an estimate of the ohmic resistance of the electrolytic solution in use. Given a solution of 2380[ppm] sodium bicarbonate solution in 25 ◦ C water, the conductance is 2071  µS cm  [1]. The area of the electrode plates is 120[cm2 ]. Given these known quantities,  some assumptions have to be made regarding the solution: • Volumetric gas content of 15% (visual estimation from experiment). • Bubble size of d = 100[µm]. • Bubbles on plate have same effect as cylinders of L = d and equal radius with regards to resistance due to area displacement. Resistance due to fluid alone: R=  L 9.52[mm] = σA (12000[mm2 ])(2380 7  µS cm  )  = 3.34Ω  (3)  Compensating for bubbles in fluid from [11]: R∗ =  R = 4.44Ω 0.75  (4)  Resistance due to bubbles on surface of plates reducing the area of the conductor:  R=  L 50[µm] = 2 σA (12000[mm ] · 0.85)(2380  µS cm  )  = 0.0438Ω  (5)  for two plates, resulting in a total of 3.37Ω. It is clear from this that the increase in ohmic resistance due to the bubbles on the electrode reducing the immersed area is essentially negligible. As the bulk bubble density is essentially constant with PWM of frequencies over 0.5[Hz], PWM would not increase efficiency in this manner. Even if the surface bubble density is 90%, a preposterously high value, the resistance contribution is only 0.070Ω. However, this fact does not discount the bubble overpotential, a different effect that acts through a much more complex mechanism [4]. Furthermore, at bubble densities over 30%, the fluid sheet becomes unstable [11] and conduction becomes sporadic, and PWM could play a role in maintaining this stability. Though there is no existing theory to quantitatively predict the effects of PWM, one can qualitatively predict its effects based on knowledge of the macroscopic reaction behaviour. After the current is switched on, bubbles begin to form, which decreases the immersed area of the plates and creates bubble overpotential. The concentration of the electrolyte at the plates also decreases, which accordingly decreases the efficiency of the reaction. Ostensibly, if the current can be turned off when the reaction efficiency decreases significantly, then one could improved the reaction efficiency by simply avoiding the inefficient part of the reaction. The degree to which the efficiency changes between the transient and steady state is unknown. An attempt was made to investigate this behaviour, but it was not possible to measure the current flowing through the cell with sufficient accuracy due to noise, ringing in the power circuits and the lack of a current sensor with sufficient sampling rate.  8  III. A.  EXPERIMENTATION Experimental Setup and Methods  The experimentation setup for the electrolytic cell was two stainless steel plates free from any visual signs of deposits or corrosion mounted inside two polycarbonate plates. The plates were separated with a 38 [in] rubber gasket. After water passed through the cell, it entered a separation tank at atmospheric pressure. Liquid water containing no bubbles by visual inspection returned into a tank to restart the cycle. Oxyhydrogen gas, now liberated from the water, flowed into a Honeywell AWM3100V 200[sccm] gas flow measurement device. As the AWM3100V is designed for standard air, the voltage supply was increased to 15[V ] to compensate for the high specific heat of hydrogen, as recommended by the Honeywell data sheets. The water flow was kept the same between all experiments. Supplying power to the cell proved more difficult than anticipated because of the high current demands of the cell, combined with the requirement of sustaining a high switching frequency. For low power testing, a lab bench power supply was used, and for pulse width modulation testing, four 15.5[V ] lithium ion batteries in parallel were used. The pulse width modulation was performed by a MOSFET based circuit to control the power through the electrolytic cell. The pulse width modulation signal was provided by a function generator and fed into the power control circuit in presented in Appendix A. The electrolytic cell solution was prepared as 2.10[L] of tap water with 50.0[g] of sodium bicarbonate resulting in a concentration of 2380[ppm] or 0.283M . According to a datasheet from a manufacturer of liquid conductance measurement devices Aquarius Technology [1], this solution has a conductance of 2071  µS cm  at 25 ◦ C and this increases by 1.8% − 2.0% per  degree Celsius. The methods used in this experiment were relatively simple. For low power testing, the voltage was varied on a power supply, and the voltage and current were measured using lab bench multimeters for higher precision. The point of gas production was verified visually, being the point at which bubbles were seen to rise from the plates. For the PWM testing, the duty cycle was varied on a function generator and verified on an oscilliscope. Frequency was also varied and verified in the same fashion. The physical setup of this experiment is shown in Figure 4.  9  FIG. 4: Physical setup of experimental apparatus. The water pump and tank are in blue, the electrolytic cell in green, and the water/gas separator in red. B.  Low Power Analysis  Investigation of the cell power consumption near the threshold of gas production yields several useful results: the threshold voltage for production of gas, and the power consumption behaviour under and over the threshold voltage. The point of gas production was determined visually, and determined as the point when any bubbles were seen to rise from between the plates. The test was conducted with water flowing and stagnant conditions in the cell at 25 ◦ C with the results in presented in Figure 5. In stagnant operating conditions, the threshold voltage was observed as 2.465[V ], and the bubbles could not be accurately observed with the water flowing. This is not far from the value predicted from a reduction-oxidation table in combination with the reaction overpotenial of 2.13[V ]. This difference in values may arise from the different electrolyte (0.283[M ] sodium bicarbonate instead of 1.0[M ] sodium hydroxide used in the reduction-oxidation table). It is also possible that the difference represents the bubble overpotential and concentraion overpotential mention prior. In Figure 5, the change of slope of both scenarios at  10  FIG. 5: Low power behaviour of electrolytic cell.  this voltage corroborates the visual analysis measurement of the threshold voltage. Below this voltage, the power is dissipated purely through ohmic resistance, so the relationship between the voltage and current allows one to determine how much power will be dissipated in this manner at any power level. Just below 2.465[V ], the slope of the graph corresponds to a resistance of 2.12Ω. However, at 2.00[V ] is another slope discontinuity as seen in Figure 6, which is not explained by the theory. Below this voltage, the slope corresponds to a resistance of 26.1Ω. This may be a nonlinear effect from the fact that this is not a traditional conductor where electrons are the charge carriers, but one where ions move through the solution as the charge carriers. The difference between stagnant and flowing conditions provides some notion of how the surface effects change the efficiency of the process. Assuming that gas production is proportional to current as predicted by electrochemistry, this observation should show if PWM power would have any noticeable effect at all. The reason for this is that at stagnant conditions, all of the bubble overpotential, concentration overpotential, and other ohmic effects decrease the efficiency of the reaction. The same is not true for the flowing scenario, because the gas production is so little compared to the flow rate of the water, that the 11  FIG. 6: Closeup of two slope discontinuities.  bubbles are stripped off and solution replenished before the overpotential effects have any contribution. At higher rates of gas production in the range that is useful for practical purposes, the same is not true. While the result does not provide a qualitative measure of estimating the potential efficiency gains from using PWM, it does show that a noticeable effect of up to 10% is possible under the ideal conditions. The difficulty in attaining these conditions comes about in the flow rate of the conducting solution. To strip the bubbles from the plates and replenish the conductors in the solution during the off portion of the PWM cycle, the flow rate must be sufficiently high. However, a high flow rate strips bubbles off before they can accumulate to a macroscopic size, and at this small size they are incredibly difficult to remove from the solution. Using DC simply aggravates this problem as the gas production is continuous. Remedies to this problem are addressed in the Recommendations section of the report.  C.  Pulse Width Modulation  Using a low flow gas sensor, the gas production was measured as a function of pulse width modulation frequency and duty cycle. The voltage was maintained at 15.5[V ] by measuring the voltage of the batteries and recharging as required. As is readily apparent from Figure 7, the current setup does not provide a sufficient 12  FIG. 7: Effect of varying PWM duty cycle and frequency. Error is ±10% in gas production.  level of precision or accuracy to make claims on the effects of PWM power. The nature of the setup does not allow direct visual observation of the gas producing surfaces; however, gauging from the rate of gas production and the bubble density in the exiting fluid, it is very possible that the bubble density between the plates is such that a loss of stability in the fluid films results in a breakdown of conductivity [11]. Multiple attempts at collecting data in various setups were attempted; however, for reasons discussed in the Experimental Error section, no conclusive quantitative results could be achieved from this experiment. The trend of the results is roughly linear, which allows at least for a rough estimation of efficiency; from the Theory section, this process will create 0.890[ mL ] or 53.4[ sccm ] . Ws W From this data, the cell only achieves 2.3[ sccm ] or 4.3% efficiency in the best case. The W reason for such a large disparity from industrial efficiencies is likely material and electrolyte differences. The electrochemical differences are beyond the scope of this project, as the goals of this project are centered on relative improvements of efficiency rather than absolute efficiency. Nevertheless, this abysmal efficiency shows how compromise in materials and chemicals affect this process.  13  D.  Experimental Error  Due to the change in focus of this project, the initial setup built for a viable amount of gas production was not suitable for experimentation for several reasons. The pump used in this setup became very hot during extended use, which heated the water from 25 ◦ C to 40 ◦ C over one hour of run time with no oxyhydrogen generation. As shown in [1], temperature greatly affects the ohmic conductivity of ionic solutions, and will therefore affect power consumption accordingly. As there was no control whatsoever over temperature in this setup, this was a variable that negatively impacted the results of the experiment. The composition of the solution and electrode surfaces also varied with experimentation: as the cell was run, deposits appeared on the electrode surfaces, which has several effects. The conductivity of the solution will decrease as ions are pulled out of solution, and the oxidation-reduction potential as well as the reaction overpotential will vary in an unknown fashion. The cell itself had intrinsic design flaws with regards to experimentation as well. Using a 3 [in] 8  gasket appeared to be an insufficient separation distance to maintain a fluid film given  the experimental solution flow rate and the power put into the cell. Furthermore, the water flow was highly irregular because it entered the cell at 90 degrees to the plates through a hole cut into the plates. In addition to the variability in the setup, the measurements themselves had a poor level of accuracy. It was found that minute variations in the water level of the separation tank dramatically affected gas flow measurement results. The degree to which this occurs is shown in Figure 8. A zero gas flow has an output of 1.5[V ], which gives a sense of scale of the fluctuations. Considering that most measurements were well below 100[sccm] or 1.67[ mL ], s this fact is unsurprising. Calming the fluid surface sufficiently while maintaining effective gas separation proved difficult and could not be accomplished in the time frame of this project. The gas sensor used in this experiment, a Honeywell AWM3100V hot wire anemometer, proved extremely sensitive for flow measurement. However, converting the output voltage from the sensor into [sccm] of oxyhydrogen introduces significant error. The sensor is calibrated to standard air, and Honeywell provides a conversion factor based on the composition of the gas. Nevertheless, this is an approximation based on the specific heats of the gas. Honeywell claims a maximum of 5% accuracy as shown in the conversion diagram of Figure  14  FIG. 8: The gas sensor records large fluctuations in the gas output of the system. From a mean of roughly 2.5[V ], the voltage fluctuates over 1[V ] in some instances.  9.  FIG. 9: Honeywell Sensor Calibration Curve [6].  15  IV.  PROJECT DELIVERABLES  The bulk of this project was an experiment, which leaves the final deliverable as this report. In the process of this investigation, the support system for the cell was constructed. What follows is a list of the required hardware for this project. 1. Electrolytic cell with nine stainless steel electrodes. 2. 80[W ] indoor hot water pump, providing 15[f t.] of head. 3. Separator tank fashioned from 3[in.] polycarbonate tubing. 4. 2[L] PTFE water tank. 5. 6[f t.] of  3 8  Tygon tubing.  6. Honeywell AWM3100V 200[sccm] gas sensor. 7. Purpose built PWM circuit. All components were purchased by the Project Lab without any budgeting from myself. No ongoing work remains on this project as the conclusions show that this setup is currently inadequate for further attempts at scientific investigation.  16  V.  CONCLUSION  Initially, this project set out to construct an apparatus that would produce useful amounts of oxyhydrogen gas for use in the Project Lab. After construction of the electrolytic cell and associated support hardware, testing showed that the amount of power, and specifically the current level that is required to produce useful amounts of gas is unreasonably high for this project. Building a power supply from scratch that could handle the required power of power levels over 2000[W ] and 25[A] would demand components whose cost would have been beyond the budget of this project. As a result, midway through the term, the project shifted to a scientific investigation of possible efficiency gains in the process through pulse width modulation of the power supply. The electrolytic cell was disassembled, thoroughly cleaned of deposits and contamination from prior electrolysis attempts, and reassembled with two plates. Reducing the number of plates made the experimental setup more closely match the existing theory of conductivity on ionic solutions and two phase solutions, as well as greatly reduce the power requirements. With this setup in place, low power investigations revealed the threshold voltage for the system as well as behaviour before and after this point. The main investigation into the effects of PWM frequency and duty cycle was also performed and the results were presented, but were not sufficiently accurate to draw any conclusions or make any claims. In conclusion, the most valuable results gained from this project are the recommendations for performing this experiment for possible future projects.  17  VI.  RECOMMENDATIONS  Because this setup was not designed for experimentation in any respect, there is much room for improvement if the goal is to investigate the effects of PWM with the potential for useful scientific results. So many variables beyond control were present in the system  A.  Cell Design  Two major improvements are possible with the current setup. First, construction of a flow modifier or cell redesign should be carried out in way that forces the water flow through the cell to be much more regular and laminar. Secondly, the plate material must be changed to a material that is inert in this reaction: platinum or gold. Gold plating was not initially carried out as the initial design had nine electrode plates. However, in light of the design change to two plates for scientific experimentation, the cost of $60 per plate is not unreasonable for a project of this scale. If gas production over 30[sccm] is desired, then the separation distance between the plates should be increased to prevent a breakdown of the stable fluid film.  B.  Support Apparatus  Much uncertainty in the gas flow measurements was caused by the variations of fluid level in the separation apparatus. In order to dampen vertical motion of the flow, a constant back pressure from the water tank could solve the problem. The current tank experiences pressures fluctuating about atmospheric depending on the fluid flow rate in the system. This can be easily accomplished by building a taller tank. Temperature control is a more difficult problem. The most obvious change is replacing the pump with one that dissipates much less heat. The cell itself also produces heat due to inefficiencies in the electrolysis process. As a result, an active temperature control system using a radiator would be optimal for scientific measurement. An unvarying temperature is not strictly necessary, because a one degree change corresponds to a 1.8% − 2.0% change in conductivity. Maintaining temperature to within a degree or two would be sufficient given other inaccuracies in the system. Effectively separating the gas bubbles from the fluid in the system proves to be challenging 18  as well. In this experiment, a balance was struck between flow rate and bubble size to allow acceptable rates of gas separation. The dilemma is in the flow rate: a higher flow rate is highly desirable to reduce the bubble density as much as possible as well as maintaining uniform conditions between the plates. However, stripping the bubbles from the plates so soon after formation means the bubbles are microscopic, and the solution simply appears cloudy due to the bubble density. These bubbles are very difficult to extract using a simple separation tank as in this experiment. If one desires to experiment with higher flow rates, an active separation apparatus using ultrasonic agitation would undoubtedly improve gas liberation from the liquid..  C.  Electrochemistry  Though sodium bicarbonate is a safe and sufficiently conductive electrode, it is not ideal for experimentation. It has the possibility of interfering with the reaction because of the bicarbonate ion, which causes deposits on the cathod. A scientific apparatus should use sodium hydroxide as it does not interfere with the reaction and is the standard electrolyte used in the electrolysis of water for industrial production and scientific analysis. It should also be used at relatively high concentrations such as 1[M ], which is standard for electrochemisty.  D.  Sensor Calibration  Essentially all of the error introduced by the gas flow sensor could be calibrated out through comparison with a highly sensitive displacement measurement method, or another highly accurate gas measurement sensor. With repeated calibration measurements to oxyhydrogen gas, one could theoretically achieve a much more precise calibration curve with a low standard deviation if the Honeywell sensor proves suitable to make repeatable measurements.  19  E.  Power Supply  Supplying power for the electrolytic cell proved to be challenging, especially in attempts to use pulse width modulation. Finding a power supply capable of switching at the frequencies and power levels of this project proved impossible. Batteries are in some ways ideal for electrolysis and are commonly used by hobbyists for this purpose. With the ability to switch on and off essentially instantly, especially compared to power supplies, batteries have a distinct advantage in this respect. However, the power output from batteries cannot be easily controlled as with a power supply. Furthermore, heating of the batteries from heavy loading can be a serious side effect; lead acid batteries may boil the internal solution under heavy use, and the lithium ion batteries used in this project can explode if they are not current limited with a fuse. Ideally, a power supply capable of 30 kilohertz switching at 20[V ] and 50[A] would be used for the two plate electrolysis setup.  F.  Pulse Width Modulation  When this experiment was conducted, there was no theory to predict the outcome of pulse width modulating the power supply; as a result the experimentation was somewhat brute force by simply testing many PWM frequencies and duty cycles. While it may not be able to obtain analytical theory of the behaviour in the cell, one could perform another observation to aid in the seach for an optimal frequency and duty cycle. If one could obtain a high speed camera and microscope, one could directly observe bubble formation and judge what timescales of PWM frequency may be useful to test. Visually measuring the effects of PWM may be far more efficient than brute force testing. A second method of determining if PWM has any effect whatsoever, is the addition of a highly sensitive, high speed current sensor. A current sensor was used in this experiment; however, it could not sample nearly fast enough to obtain useful results. A current sensor that samples on the order of 100[kHz] to 300[kHz] would be useful in observing how the current changes as electrolysis takes place and gas is produced. The high sampling frequency is required because the transient response of the system is on the order of microseconds, and much may happen during the transient response.  20  VII.  APPENDIX A  FIG. 10: Pulse width modulation circuit.  21  [1] Aquarius Technologies, (2000) Retrieved from [2] J. Calvert, (2008) Retrieved from jcalvert/phys/hydrogen.htm  [3] Royal Society of Chemistry, (2003) Retrieved from [4] C. Gabrielli, F. Huet, R.P. Nogueira, Fluctuations of concentration overpotential generated at gas-evolving electrodes. Electrochimica Acta 50 3726(2005). [5] W. Griesshaber, and F. Sick ., Simulation of hydrogen-oxygen systems with PV for the selfsufficient solar house. FhG-ISE, 1991. [6] Honeywell,  Airflow Sensors AWM3000 Series Calibration Manual Retrieved from 71.pdf [7] European Commission on Hydrogen Safety, Hydrogen Fundamentals, (2007) Retrieved from Ch1 Fundamentals-version%201 0 1.pdf [8] I.H. Levin, The Journal of Industrial and Engineering Chemistry 5, 496 (1913). [9] J. Rossmeisl, Z.-W. Qu, H. Zhu, G.-J. Kroes, J.K. Norskov, Electrolysis of water on oxide surfaces. J. Electroanal. 607 83 (2007). [10] E. Viall, Gas Torch and Thermite Welding (McGraw-Hill), p. 4. [11] A.P. Vasil’ev, Journal of Engineering and Thermophysics, 39, 649(1980). [12] N.A. Kelly, T.L. Gibson, D.B. Ouwerkerk, A solar-powered, high-efficiency hydrogen fueling system using high-pressure electrolysis of water: Design and initial results. Int. J. Hydrogen Energy, 33 2747 (2008). [13] R. Ono, and T. Oda, Spark ignition of hydrogen-air mixture. J. Phys.: Conf. Ser. 142 012003  22  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items