Why understanding data is the secret key to unlocking an investable green energy future
By Gareth Brown, Chief Executive Officer, Clir Renewables
Despite the disruption from Covid-19 in the previous 18 months, projections for the solar and wind energy industries are looking strong. According to the International Energy Agency, 270 GW of renewable capacity are expected to be added in 2021 and 280 GW in 2022. While this is a hugely exciting growth trajectory, there are still some impediments to the increased deployment of both wind and solar power. As renewable energy technology has matured, governments have stepped back from underwriting its deployment through tariffs and incentives. This has created what’s termed a ‘post subsidy’ marketplace.
What this means in practice is that where governments have stepped back, private capital must now come forward. Capital will only take investment risk if it believes it will see a return and believes that its investment will be protected during the lifecycle of renewable energy facilities.
Protecting capital and driving investor returns in wind and solar has become an industry in itself. Firms like Clir Renewables have invested significantly in creating intellectual property that allows access to massive amounts of industry-wide data and combs data using the latest artificial intelligence (AI) and machine learning (ML) technologies. By applying AI and ML technology designed by the industry to massive amounts of renewable energy operational data, all project stakeholders, from owners and investors to insurers and lenders, can know where they need to make operational changes, and manage the performance and risks of assets. As our datasets increase in size and our modelling techniques advance, we find that we can better help guide the operational running of wind and solar projects for improved project returns.
For example, to manage the longevity of equipment, which helps keep renewable power prices low, manufacturers will ‘derate’ their wind turbines — that is, intentionally impede their power output. The unintended consequences of derating can reduce project power generation for extended intervals, rather than the specific time period intended. This is because derating strategies are often employed simplistically.
Avoiding pitfalls like this – and understanding the interplay of the different dynamics that affect project performance – is only possible with deep and granular data. Data needs to be leveraged holistically, using not just the data from individual wind turbines, but from the entire project site.
If wind power generation is to be intelligently managed for maximum returns, then it’s also important to understand what environmental factors are also at play – the effect of nearby vegetation, for example. To provide this understanding, we not only need to look at equipment, but we need to measure and monitor multiple streams of data.
Similarly, reducing the financial risk of wind and solar projects can be achieved through in-depth analysis of their components. Too often there is a reticence among asset managers to embrace data at a deep level, invest further time and resources into its understanding, and then implement changes to asset management programmes.
This hesitation may come from a perspective that ‘current performance is good enough’. Yet, with the shift to post-subsidy markets, investors are recognizing the better margins, and reduced costs of insurance, financing, and operations, available through an increase in power generation – and subsequently energy sales – of just one to two percent.
Similarly, across the solar industry, an assumption that greatly impedes optimal results is the idea that solar energy is simple. Although solar photovoltaic plants may not have as many moving parts as a wind turbine, this does not mean that data collection – which is essential to understanding asset health – is not challenging.
Gathering solar data is made complicated through the sheer number of original equipment manufacturers (OEMs) existing in the space. While wind energy has a smaller number of OEMs – owing to the higher costs of wind turbine development and production – solar has hundreds of different manufacturers. This creates data choas, as data from their individual components is labelled differently.
Additionally, the unpredictability of solar resources, the lack of standardization across components, plant topography and the quality of project data can all make it difficult to gain a clear understanding of the project’s health and performance.
Attempting to translate all of this information into an understanding of asset performance, and then contextualising it with the other elements needed, eludes most owners and asset managers. However, with an AI-driven platform, the time needed to translate and analyze data can be reduced to hours rather than weeks, as the many streams of data in different OEM ‘languages’ can be translated into a coherent thread that identifies specific actions necessary to improve performance.
As the transition to renewable energy progresses, and we see more projects going live, it is important that the knowledge around data continues to progress. This will allow projects, and by extension, their return to investors, reach their full potential. At Clir, we have access to over 100 GW of operational renewable energy projects worldwide – helping asset owners, investors and energy companies enhance their returns, ensure the smooth and efficient running of projects, and fulfil their obligations to low carbon power supplies.
With AI-driven insights, renewable energy can easily meet its future growth trajectory – and provide investors with stable long term returns while doing so.