Tuesday, May 10, 2011

Lab Testing

Laboratory Testing: Methods and Approach

The goal in the lab was to identify and optimize a successful way to produce biodiesel given the project constraints. Using the nickel-oxide coated microreactors and having selected methanol and soy oil as the reactants, the group manipulated other variables based on research and software simulations to identify a successful process. Namely, those variables manipulated specifically in the lab were the temperature, the residence time, and the molar ratio of methanol to soy oil.

The lab set up in the first semester consisted of a syringe pump, heating tape, variac, multimeter, thermistor, micro-reactor, and test tube. A picture of this set-up is shown in Appendix B. The syringe pump was a good positive displacement pump with two syringes (one for each reactant) that could pump each fluid separately until they mixed in the microreactor. After leaving the microreactor, the product flowed into the test tube for collection and testing. The reaction was heated by placing the microreactor on a heating tape whose temperature was controlled by the variac and measured by the multimeter attached to a thermistor. In the second semester, the set-up stayed the same except all the components used for heating the microreactor were switched out for a hot water bath. The microreactors were submerged in heated water and the heat was both controlled and measured by the controls on the hot water bath. Using the hot water bath in the lab was crucial because it was a safer means of heating the reaction. This meant that the process could be left unsupervised, allowing many more tests to be run throughout the semester, despite the slow flow rates of the reactants. Appendix B also shows a picture of this set-up. In the picture, solar power is being used to power the pump and the hot water bath; this shows the success of the solar power goals of the project; however, for research purposes in the lab, no further investigation of the solar power was needed to continue to run experiments, and hence, solar energy will be discussed in more detail later in the paper.

The different variables in the laboratory experiments were controlled by the temperature control and the syringe pump. The residence time was an average time the fluid would be in the reactor given the dimensions of the reactor and the sum of the two flow rates being controlled by the pump. The molar ratio was controlled by the chemistry of the reaction, and could be adjusted if more methanol was desired to drive the reaction. As will be shown later, an excel spreadsheet was made that provides these calculations; by inputting the desired residence time and molar ratio to be tested, the respective flow rate for each reactant was given. Specifically, this spreadsheet analysis will be discussed in detail in the Theory for Design portion of the paper.

As different tests were run at varying speeds, temperatures, and ratios, the data from the software simulations was used to try to gain a better understanding of the nature of the flow and of the reaction within the reactor. By changing the variables in the simulation, different variable combinations were developed that seemed to provide better conditions for the reaction to occur, and then those values would be tested in the lab.

The plan was to then optimize the process and create a final design. This involved finding the optimal temperature for the reaction and finding the most energy efficient way to heat the microreactors. This also involved a balance of flow rate speed and the percentage of completion of the reaction. The faster the flow rate, the better, but it had to be slow enough to allow for the reaction to occur. Optimization also involved a study of what percent of biodiesel produced would be desired in order to make use of the faster initial reaction rates. The group was not actually able to get to the optimization step within the lab, however, because it was never successful in producing biodiesel. Therefore, for the design requirements of the project, the group was forced to move forward with an optimized design based only on theoretical computer simulations, which, unfortunately, have not yet been proven to be consistent with the real-life lab results.

Table 1 below shows the details of the fourteen experiments ran over the course of the year in sequential order. The group initially started out with shorter residence times and an abundance of methanol. Later, it was decided that the reaction was not being given enough time, so the residence times were then greatly lengthened which was possible by the usage of the hot water bath. Using the software, the group also honed in on what was simulated to be a better molar ratio to create flow properties that were best for the reaction. As tests continued to be unsuccessful, the next step, after doing an analysis of the increased internal pressure in the reactors, was to gradually increase the temperature. Originally the group wanted to stay below the boiling point of methanol at atmospheric pressure but gradually became more daring at risking boiling some methanol off in order to go to higher temperatures to try to initiate the reaction. For the final experiment performed, the group sent the results away for analysis by Nuclear Magnetic Resonance in order to be sure that the Thin Layer Chromatography (TLC) analysis being done in the lab was not missing some amount of biodiesel.

Sunday, May 8, 2011

Final Design

The final design of the Green Machine consists of several key components: a pump, a solar array, tubing, microreactors, an immersion heater and a collection bin. The system consists of 750 microreactors in parallel which will produce exactly 1 gallon of biodiesel in a 24 hour period as specified. Each microreactor will be 7 cm long and the reactants will have a residence time of 30 seconds. The immersion heater contains racks on different levels. The microreactors will be stacked 30 high and 25 across within the immersion heater. The heater itself will be maintained at a uniform temperature of 80˚C, requiring 106.1 kW of power. The dimensions of the immersion heater are 120 in. by 100 in. by 8 in. The pump used will be a Harvard Pump 33 Dual Syringe Pump because it operates well at low flow rates and can handle the pressure drop within each microreactor of 1156 Pa, which is less than the specified maximum of 1,300 Pa [ref 8]. This pump can operate at 45 W which is more than enough to overcome these pressure drops. The system will pump the reactants at a molar ratio of 25:1 for methanol to soy oil, resulting in a total flow rate of 0.0035 mL/min. The flow rates of methanol and soy oil are 0.0027 mL/min and 0.00084 mL/min respectively. Lastly, the solar array will be 108 m2 in area, making it capable of providing 108 kW of power to run both the pump and the immersion heater.

The image below shows the final design of the Green Machine. Although this is just a basic set up, it clearly takes into account all of the necessary calculations and specifications as discussed above. In this design, the pump rests on a column, feeds tubing into the immersion heater, pumps the fluids through the microreactors, and then more tubing empties the biodiesel into a collection bin. The solar array is mounted on the column upon which the pump rests. Although it cannot be seen in the schematic, the column is at an upward angle so that the array is exposed more directly to the sun. Additionally, the system is to be constructed so that the solar array is at a southerly exposure point where it will receive the most sunlight (provided it is built in the Maryland area). This design sufficiently meets all performance specifications laid out earlier.



In a future post, we'll discuss exploring the reasons why the microreactors failed to yield biodiesel.

Tuesday, April 26, 2011

Wrapping up for the Semester

As the semester winds down we have a number of issues to comment on.

Thus far, we have been unsuccessful in producing biodiesel using the microreactor technology. Currently we have sent a sample to a testing facility; we believe that, since TLC is a relatively imperfect testing method, that we may have generated a percentage of biodiesel in our product which we are unable to detect.

We have been conducting exhaustive tests, as stated previously, on the variables relevant to our experiment in a process of elimination to discern where our process is wrong. As the batch tests conducted by previous experiment have shown, when these substances (soy oil and methanol) react in the presence of nickel oxide they produce biodiesel. One solution that seems extremely probable (and difficult to test) is that there is simply not enough surface area in the microreactor with which the reactants can come into contact with. However, we are ensuring that there can be no other possibility of failure.

On a more positive note, the solar aspect of the project has been shown to successfully run the experiment. Based on calculations of power and the picture of the wave generated by the solar cell, the solar panel we have is actually larger than needed to run the Green Machine. This is favorable as there is a non-negligible power loss when converting the DC current of the solar cell into the AC current needed by the pump and heater to work at the moment.

As we prepare for the presentation for the IAB, we are ensuring that no stone is left unturned, even considering how the Green Machine would work and be designed should the microreactors yield biodiesel.

Monday, March 14, 2011

New Semester: Hitting the Ground Running

After the long winter break (maybe not long enough), we entered January with a strong drive to start moving forward as quickly as possible with the Green Machine. Beginning with a re-division of labor and a re-ordering of priorities, the team started working almost exclusively on the successful production of biodiesel.

Immediately during the last couple weeks of January the team ran exhaustive lab tests. Modifying temperature, length of time to run the test, temperature of the reaction, the tests were run while changing these conditions individually so as to not confuse results. Meanwhile, Tim continued to enhance the software modeling and I sealed up the solar power portion of the experiment for the time being to as to devote more resources to the actual experimentation.

The results so far have not been good, but thus is the nature of research. We have yet to produce biodiesel with the microreactor technology, but this does not sound the final bell for the project yet. Aside from running a slug test (allowing the materials a set amount of time to sit in the microreactor channel and mix with the catalyst before removing the material(s) and substituting fresh reactants), we are currently performing root failure analysis to determine what exactly is causing the reaction to fail or, rather, what is missing from the reaction that is not allowing it to occur. One facet we are fairly certain of is this: the catalyst, nickel-oxide, definitely works. All that is left is to exhaust the other possible solutions.

What lies ahead for the rest of the semester is not cemented and depends on whether we can actually generate biodiesel. Successful generation would allow us to turn our attention to other facets of the project, perhaps even approaching a final design concept for a complete generator. Failure to generate biodiesel with the microreactor technology opens up new research options - isolating the components of the experiment that are lacking, suggesting new directions to move in for the catalyst (which is still believed to be superior to current biodiesel generation methods). Both scenarios will entail a great deal of economic analysis, scaling research, and more modeling.

Monday, December 13, 2010

Plan for Next Semester



Above is a detailed layout of our plan for next semester. The tasks are organized in order of importance. We hope to have our design completed construct by the beginning of March. We will then begin to perform test in the laboratory on our design.

Enviornmental Benefits



Biodiesel fuel reduces carbon dioxide emissions because the CO2 is cycled. According to the National Renewable Energy Laboratory, and as shown in Figure 1, the use of 100% biodiesel fuels can reduce CO2 emissions by 78.45% compared to petroleum biodiesel.

Serpentine and T-shaped Microreactors