All Energy 2014 – Pelamis cost of energy study

Background

In the last session of All Energy 2014, Jacob Ahlqvist presented key findings from Pelamis's recently completed cost reduction project commissioned by the ETI. The project goals were to:

1. maximise returns from the two P2 machines presently being tested at Orkney
2. drive down cost of energy
3. deliver the next design iteration, the P2e, as quickly as possible

Jacob noted that the P2e would 'address the short-comings of the P2', i.e. apply lessons learnt from testing at Orkney to design changes targeted at improving the cost of energy. The final project report was reviewed by independent consultants Black and Veatch, and approved by the ETI review panel. The report itself is not publicly available, but I've heard there are plans to disseminate the results. In the meanwhile, here is what I jotted down during the talk.


Methodology

Jacob's team produced cost curves for Pelamis: these showed how learning by doing would result in a reduction of cost of energy over time. They built cost models, using existing methods by the ETI/UKERC and Carbon Trust as a guide. The method differentiated between conceptual, manufacturing, and operational learning, which take place at different stages in the development process. The design was disassembled into sub-systems, and again into components. 'Learning rates' (rate of cost reduction with installed capacity) were assumed for each component, taking into account supply chain consultations.

The study included an innovation program. This began with the identification of areas that could benefit from innovations. Areas with the most potential were selected for in-depth study. These were reviewed and ranked according to models of performance, reliability, and cost of energy. In-house financial (internal rate of return) models were adapted to output cost of energy.



Results

Cost curves showed the cost of energy plotted against time (Note: usually such curves are plotted against installed capacity, so this study must have assumed a given installation rate. Details of the installation rate were not given; typically the early uptake of new technologies is exponential). The curves showed high initial costs in 2018: 280 £/MWh (with uncertainty bars from 180-380 £/MWh). The model suggests that by 2023 learning will reduce costs to 150 £/MWh, and by 2028 to 100 £/MWh.

The 'cost to commercialise' was also investigated. The development costs required to reach a target cost of energy were plotted against time. Two cases were compared: one where cost reductions were made only through learning (experience and economies of scale), and another where R&D was conducted at an early stage, e.g. the cost-reducing innovations identified by the innovation program that was a part of this study. It was shown that the cumulative development costs were lower when cost reductions were introduced at an early stage. Jacob noted that this demonstrated the importance of investing in 'the right machines'.


Discussion 

During the audience question session, I asked Jacob what 'the right machines' were, right now. His answer, 'not arrays', was extraordinary.

Some background information is required at this point. Technology development levels (TRLs) are used to describe the progress of commercialisation of marine renewable technologies. According to this metric, the next step for Pelamis should be a pre-commercial array; this would take the technology to the final step of 'TRL 9'. There is plenty of public support (both financial and political) for such demonstration arrays, yet to date, only tidal stream has tapped into this. However, there is EU funding to build a wave array in Ireland (ESB's Westwave project). The Irish utility ESB is still deciding which wave devices will be installed, but as Pelamis has the most operational experience, it is on the shortlist. All this suggests that it would make commercial sense for Pelamis to pursue a pre-commercial array as the next step in its development path.

It is not surprising that the assumed development path suggested by the technology development levels (TRLs) has been called into question. Neither is it surprising to observe that the bulk of the funding available (supporting first arrays) is for work that does not support the most cost-effective development path. Indeed, these concerns have been increasingly acknowledged in both academia (Jochem Weber's work on the readiness/performance matrix) and in industry (ESB's inclusion of performance criteria into the TRL definitions). What is surprising, is to hear this stated publicly by a developer, when the funding for their preferred development path is so limited.

Jacob added that what was needed now was machines which would give a better knowledge of machine operation, as this is influenced by many facets of the design, and which would investigate innovations to improve cost of energy. Another audience member asked Jacob where he would bet on cost reductions. Jacob answered that yield, theoretical limits, control algorithms, capital costs, complexity, and amount and type of materials, were all promising areas.



Conclusions

The ETI cost reduction program has identified several innovations that could reduce the time and cost to make Pelamis competitive with other offshore renewables. However, the present funding environment, which supports a one-way fast-track from concept to demonstration array, is not suited to incorporating these innovations. Pelamis will need to work hard to gain political support for funding of its iterative design process. Other marine renewables developers suffer the same commercial pressures, and may also benefit from such support.

Acknowedgements

Thank you to Jacob Ahlqvist and Deborah Smith of Pelamis for ensuring I gave a fair account of this seminar.

Image credits

'Tiananmen-ducks': http://en.wikipedia.org/wiki/Rubber_Duck_(sculpture)