Stage 2 is hard. Stage 1 of the Wave Energy Prize took some time, but only required describing our wave energy converter (WEC) in words and pictures. In Stage 2, we have to prove the concept works, both physically with the 1/50th scale model and computationally with numerical modeling. And the physical and computational results have to match. All of which is doable, and constitutes a normal part of the design development process.
What make is hard is that there are only two of us and we only have three months.
In Stage 2, we are responsible for submitting 3 key elements:
To understand Stage 2, it helps to look at how the Wave Energy Prize judges assess it. The table above is from the Wave Energy Prize Rules, Appendix E. I think their method of assessment demonstrates a very good overall understanding of WEC design development.
In the first data column of the table, notice that 15% of our score is based on the power performance (Capture Width is a measure of power absorption) of our 1/50th scale model. This may seem low: isn't the point of Stage 2 to prove our concept with a scale model of our device? Not exactly, in this stage, the numerical modeling and overall concept are also very important.
25% of the score is the correlation of the experimental test results with the predictions from the numerical model. Here, emphasis is placed on the numerical model as well as the scale model. In general, getting a match between numerical modeling and physical results is a standard and important aspect in concept design development. Typically, this is done by collecting physical test data on device motions and performance, and tuning one's numerical model to match these results, which is done by adjusting parameters that model viscous drag. Once a good match has been made, the numerical model can be run to make predictions of performance for cases outside those testing in the wave tank, i.e. extrapolating the results.
The challenge, and this is pretty major, of this Stage of the Wave Energy Prize is that the numerical modeling results are submitted blind; i.e. they are submitted before the tank tests are actually conducted, and so contestants don't have a chance to tune their numerical models. This means that to really be successful, contestants need to actually conduct independent wave tank tests on their physical model prior to shipping it to the Prize tank, and use their own results to tune their numerical model.
We are fortunate to have access to the University of Edinburgh Curved Wave Tank where we will test our model. However, contestants who cannot test in a tank will be at a disadvantage. Throughout the competition, the Wave Energy Prize has made it clear that they wanted the competition to be accessible to anyone, even those not in the field of wave energy to attract outside thinking. However, in my opinion, it would be very difficult for someone without a background in wave energy or another marine field to have the knowledge and skill set to compete effectively.
This also brings us back to the challenge of Stage 2: we need to build a 1/50th scale model, create a numerical model, test the 1/50th scale model, and correlate the results in 3 months. (We found out that we made it to Stage 2 on August 15, and our model needs to be at the tank by November 23 and our complete submission needs to be in by November 30.) If I were designing a project like this, I would give it at least 6 months, and careful design and testing regime could take a year.
The next scoring criterion is 30% for the revised Technical Submission. I believe this is to give the judges a chance to compare the concepts in other categories besides power performance. E.g. the judges can say: "ok, this device performs really well, but it will be extremely expensive to install and maintain."
Finally, 30% of the score is the estimated performance of the WEC in the final competition stage, the 1/20th scale test. Here the performance is the ACE (Average Climate Capture Width per Characteristic Capital Expenditure), which is an estimator of the cost of energy produced by the device, and takes into account the power performance as well as the amount and cost of structural material used in the device.
The power performance is partially based on extrapolating of the numerical modeling results. For the numerical modeling, we can submit 2 sets of results, one to compare to the 1/50th scale testing, and one with a more complex control strategy that we would employ in the 1/20th scale testing. Unfortunately for us, we don't have a more complex control strategy at this time, and so cannot take advantage of the dual results submission.
The wording of the final scoring category is vague enough to possibly include a subjective assessment of the performance, which seems fair. This gives the judges a chance to consider how realistic a device's control strategy and numerical results are.
Stage 2 is a lot of work in short amount of time. Our design and testing process is being highly accelerated with critical deadlines. But we're not complaining; it's exciting and it's part of the challenge.
Stay up to date
These posts are intended to help people understand what it is we're doing.