SR article

Preliminary Experiments

Our initial goal for this project was to determine the optimal process parameters by conducting experiments. When we first arrived in India, we spent time making an experimentation plan, in which we determined what experiments we wanted to do, and a schedule of when we were going to conduct which experiment.

However, as mentioned before, we never got to put this plan into action since there were always new technical difficulties setting us back. We did a few experiments, however these weren’t formal experiments following our schedule. Since we had no rotameter at this point, we didn’t know what flow rate we were using, which is crucial for the results and the interpretation. Hence, these experiments were done in a rough manner, and to us, they served more as a way to get more familiar with the process than actually contributing to the research in a scientific way.

Therefore we decided to call these experiments “preliminary experiments” that had to be done before doing the experiments that would provide precise numbers and results.



Experiment 1 – 18.12.2016


  • Methods (Parameters)

We used 8.4 kg of grass. We kept the temperature between 110 and 120 °C and the pressure around 1.8 bar. In order to keep these parameters constant, we often had to adjust the flow rate of steam into as well as out of the unit. We ran the distillation for 2h30min.

  • Results & Observations

We obtained 6mL of oil. There were some leakages when reaching a pressure of about 2 bars. The grass was still quite green when we removed it.

  • Conclusion

We had a low oil yield because we used little water (we estimated about 10L in 2,5h), which is not compatible to the steam to grass ratio calculated (8L in 1h). In other words, the flow rate was too low.                            



Experiment 2 – 19.12.2016

  • Methods (Parameters)

We used 8 kg of grass. We kept the temperature around 120 °C and the pressure around 1.8 bar. In order to keep these parameters constant, we often had to adjust the flow rate of steam into as well as out of the unit. We used a higher flow rate than in the last experiment, meaning that we opened the steam tap more. We ran the experiment for 1h.

  • Results

We obtained no oil. This was due to the fact that there was only steam coming out at the outlet and no liquid.

  • Conclusion

The coil is too short. When using a higher flow rate, the steam that passes through the coil is not cooled down enough for it to condense before it enters the separator. The oil was thus released into the air along with the steam and this is why we lost all the oil produced.



Experiment 3 – 22.12.2016

  • Methods (Parameters)

Since the flow rate was too low in the first experiment and too high in the second, we decided to use a low flow rate that allows the water to condense inside the coil, but also to increase the water to grass ratio by using less lemongrass. We used 4.1 kg of lemongrass and kept the temperature between 115°C and 120°C and the pressure around 1.8 bar. We ran the distillation for 1.5 h.

  • Result

We were able to distil 1.5 mL of oil.

  • Conclusion

We concluded that a smaller amount of lemongrass results in a smaller yield, even if it increases the water to grass ratio. It is better to choose larger quantities and packing densities of grass.




Overall Conclusion

As mentioned before, these experiments did not deliver much exact scientific knowledge. However from conducting these three experiments, done in different conditions (similar temperatures but varying flow rates and grass weight), we were able to note a very important design flaw: The coil in which the steam-oil mixture should condensate is too short. The function of the coil is to turn the steam-oil mixture into a condensate, so that the oil and water can be separated from each other in the next step. We concluded that this coil was too short to fulfil this purpose and that it had to be replaced.

This way, these preliminary experiments did deliver new knowledge about the prototype. Also, they served in making us more familiar with the process. Through these experiments, we knew much better how to handle the unit and this helped a lot when writing the operational guide, which will be further discussed in the next section.