Based on our completed testing results, the following changes to our specifications and next prototype involve:
Focus on the transient nature of the system– Forego calculations that rely on steady state assumptions. The heat exchanger system relied on calculations for effectiveness and heat transfer that involved steady state assumptions. When testing, we realized these assumptions did not make sense. Our data is best shown over time which makes it difficult for us to compare to theoretical predictions. It is important for future teams to venture into unknown calculation territory, rather than sticking to equations with steady state/strict assumptions from previous coursework.
Clarify most useful metrics– Some metrics like pump noise for our heat exchanger were not a consideration in the later stages of our design. Prior to building, focus on what metrics are most useful for presentation. These would be values like roundtrip efficiencies, heat transfer quantities, and system noise. Enhance metrics that relate to larger scale systems like what the ChemE team worked on.
Organize Design Selections- Have a spreadsheet and use (like thermal fluids design report) to assess different tubing diameters, materials, tubing length and their impact on design (heat transfer). Materials selection was based on one metric like price and it should be selected from a variety of different metrics.
Create Clear Workflow- We would suggest to future groups to start working with the turbine/expansion system and decide on the best way to demonstrate power output. Also select an appropriate wattage for power output so that your design selections for the other subsystems are narrowed. We started with an idea to power an electric vehicle, and realized our budget would not support this. By selecting an appropriate power output first we would have saved time, allowing us to spend more time on data analysis.
Further Divide Subsystems- Initially, we assumed the heat exchanger system to be a single subsystem. In reality, we had three separate heat exchangers operating for the entire system. Future teams should understand the need to further divide up subsystems so that theoretical predictions are done more accurately.
IDA- At time of project completion, our Arduino is able to correctly gather temperature data and we have a MATLab file that simulates expected efficiencies and works in a general output print format. Improving the instrumentation and data acquisition process would allow for bridging between Arduino and MATLab to automate the process of turning temperature, pressure, and flow recordings into process diagrams and charts that clearly quantify the actual and expected work input/output, energy transfer, and efficiencies.
Safety- For the demo, we used lightbulbs and their luminosity to illustrate the intensity of the discharge and how that fluctuates over time. If the future team continues with this idea, it would be safer to install a removable covering around the light bulbs since they tend to get really hot. Furthermore, touching the bulbs is ill-advised since the oil and sweat from our hand could react with the glass surface to form decomposition compounds that absorb heat over time and thereby reduce the lifespan of the bulb.
Currently, the drive belt is compartmentalized within a cage – where debris and thin objects could still get caught unintentionally. For instance, long hair could slip between the caging ribs and get tangled in the belt drive wheel and be ripped off a person. Clearly this poses a safety risk. It ay be helpful to add a plastic casing around the metal caging to remove the gaps all together while still providing visibility.
Sustainability- For future teams that want to work on this project, it may be helpful to scale the project and eventually source the power from renewable technology such as solar, wind, and geothermal power rather than a fossil fuel ran grid.
Scalability– If implemented into Lafayette Campus, our CAES design could be scaled to take up the space of a utility room at the base of College Hill. Not only would scale improve the efficiency of various system components, it would also provide for a lot more power that could eventually go on to power electric vehicles and the like.
Recommendations for the subsystems:
Compressor- There is room to scale if we want CAES to be economically viable. At the moment, our system is unable to output sustained power for more than 15-20 minutes which is a problem if CAES is to become an energy solution. If the compressor and pressure vessel goes to take up an entire room, this could allow for multi-hour power provision upon discharge which is much more useful. When scaling the pressure vessel and compressor, it would be necessary to change the means of compression. At the moment, there is a drive belt and a two stage compression process. Industrial compressors are likely to implement rotary screws and three/four stage compression processes to improve efficiency and air flow.
Heat Exchanger- At the moment, due to lack of scale, we do not anticipate the water in the thermal storage to reach a temperature that we should be concerned about (~ 398 K). With scale, the material of the thermal storage bin should be more insulative or be wrapped in insulating material. Adhesive insulation pads can reduce air space and help with that. In addition, the tubing in which water flows between the thermal storage and the intercolling component can be better tied down – both for aesthetic and organizational reasons. A tag should be added to the piping and tubing to indicate which part of the process flow diagram it implements and make it easier to do maintenance in the future if the original team is no longer able to.
Expander Generator- At the moment we are turning electrical energy from the campus grid into pneumatic energy (in the form of compressed air) into rotational energy at the generator and finally into electrical energy. In the process, all the energy conversions have high transactional costs and significant inefficiencies. The generator has little room for efficiency gains since it is currently at 90-93% efficiency. On the other hand, the air motor expander could be improved upon. When the expander is scaled up, the relative impact of friction is reduced and the efficiency will increase. Our current air motor was made for manufacturing implementations and not optimized for efficiency.