Growing Pain #2: Tackling bad data practices in the solar sector

A key goal of the Growing Pains series is to highlight the role that reliable solar resource data plays in ensuring the success of large-scale solar projects. The previous article demonstrated the impact of using accurate data during the project design phase.

However, within the solar sector, there are often apparent commercial incentives to use certain data sets over others, regardless of their accuracy and reliability. This encourages practices such as ‘data shopping’, or selecting datasets for short-term gain, due to a lack of awareness and transparency within the sector.

With multiple free and paid options available to developers when selecting a dataset, making an informed and impartial decision is crucial.

The critical question to ask is whether a chosen data source is reliable. Following best practice and improving transparency between stakeholders will ultimately ensure higher profits overall while rewarding those who invest in understanding the true potential of their large-scale projects.

This instalment of Growing Pains outlines the challenges created by data shopping and the ‘paradox of choice’. It explains Solargis’ recommendations for tackling these issues – and lays out why taking steps to improve transparency now will help industry players deliver healthier, more sustainable growth over the coming decades.

The risk of bad data practices in the solar sector

The reasons that unreliable datasets are used in the solar sector fall under two main categories: ‘data shopping’ and the ‘paradox of choice’. Individuals and organisations may not be aware that they are taking these approaches, which is why raising awareness on all sides is essential.

Challenge 1: Data shopping

When securing project finance or selling a project, the highest energy yield prediction will ensure the most favourable financial terms. To produce the largest numbers, the temptation is to choose the dataset that illustrates the highest levels of solar resource, ‘shopping around’ for the data source that suits a particular short-term financial objective.

Outlined below are two examples of how data shopping can occur.

1. Mandating a preferred data source to Independent Technical Advisors

The unbiased insights that Independent Technical Advisors / Engineers can deliver add value to the development process. If it becomes common practice to require advisors to use a particular dataset, it undermines their independence. Advisors are then incentivised to produce the results that their clients want. There are parallels to the financial industry in 2008, where credit agencies gave more favourable ratings to lenders to make themselves more attractive. The project evaluation should be based on a choice of datasets that can be objectively evaluated as reliable.

2. Discarding inconvenient measurements

It is best practice to measure the solar resource on site, for a year or longer, to reduce the uncertainty of satellite-based solar model outputs with accurate ground measurements. If solar model data adapted for the local conditions proves to be lower, it may lead to a temptation to disregard the more reliable, validated results in favour of the original estimate.

Not following transparent processes could have short term benefits for some stakeholders. However, if one stakeholder benefits at the expense of others, this affects the sustainable growth of the industry.

Crucially, as the trend for developers to operate their own projects increases, it is in their own long-term interest to use consistent, validated and accurate data throughout the development process.

Data shopping therefore leads to the following risks:

• Long-run financial underperformance of projects – Inflated energy yield assessments can have serious long-term financial consequences. A 1% decrease in asset performance means lost revenue of EUR 1,000/MW/year, adding up to millions over a typical 100MW project lifespan.
• A loss of investor trust – DNV have highlighted that underperformance is widespread and, on average, projects underperform by 5%[1]. If this continues, investors will start to view solar as an unreliable asset and project finance terms will worsen.

To tackle these risks, Solargis recommends:

• Improving transparency – Make the processes that influence data selection easily accessible to stakeholders, including justification for using a particular data source in financial reports.
• Asking the right questions – Financiers, developers, and independent engineers can each scrutinise how data underpinning crucial strategic and operational decisions during a large-scale solar project’s lifecycle is calculated and selected, leading to better outcomes for all.

Challenge 2: The paradox of choice

The second key data-related practice that increases risk during the development phase and beyond is often unintentional. It can nevertheless result in underperforming projects and a lack of investor trust.

The commonly used photovoltaic simulation software, PVSyst, allows users to import solar data from over 15 sources, including a mixture of free and commercial options. Without going through a robust evaluation process, unreliable datasets can easily slip through, as the overwhelming choice creates the impression that all data is equal. This is the ‘paradox of choice’.

Selecting an unreliable dataset may lead to the following risks:

• Overestimating project production – Data shopping invariably leads to overestimation. Unreliable data can also expose large-scale asset developers to this risk, leading to a loss of trust and an underperforming project.
• Underestimating project production ­– Failing to recognise the true potential of a site can lead to less favourable project finance terms, and also reduced final sale prices.
• Technical risks – the lifespan of key components such as inverters is shortened by performance under extreme conditions, increasing maintenance costs and reducing overall project life.

To tackle these risks, Solargis recommends:

• Building awareness – Developers and other stakeholders will need to share an understanding of why specific data choices are made.
• Standardising data selection – Assessing how the data is calculated and validated, how the validation has been done – and by who – ensures that unreliable datasets can be identified.
• Applying the ‘MASTER’ process – Solargis has produced an e-book outlining the 6 properties of a reliable solar dataset, from solar specific design to temporal and spatial resolution. Download at this link to learn more.

Top three tips for selecting a reliable dataset for a large-scale solar project

Below are three general guidelines to help developers select the most accurate dataset for large-scale assets.

1. Opt for a validated data source that has consistently proven to be reliable in the specific region the site is located. For large-scale projects, long data history and high temporal and spatial granularity are important.
2. When selecting a data provider, don’t default to the lowest priced data. Invest in a provider who can be a trusted long-term partner and give value-adding support for large-scale assets.
3. Think beyond the immediate development requirements. Using a consistent and accurate dataset at all stages of a project lifecycle, including regular asset monitoring and forecasting, ensures the optimal outcomes.

Solargis has expanded on this topic in our article, ‘What to expect from your solar data provider’. Learn more here.

===

Improving transparency around data practices in the solar sector is an industry-wide effort. Solar project developers have a vital role to play. Whether that is critically evaluating their own data selection processes or providing valuable scrutiny to other stakeholders, the benefits to the sector range from tangible financial returns to an atmosphere of mutual trust.

As projects increase in scale and the stakes are raised, both of these benefits will be crucial to ensure long-term success.

Get in touch to find out more about how Solargis can advise on eliminating bad data practices and increasing transparency for large-scale solar projects: https://solargis.com/about-us/contact

[1] https://static1.squarespace.com/static/5b4e34d1f2e6b166c33dc4f1/t/5ee91426b832f31dd0d8c26b/1592333358437/Solar+Risk+Assessment+2020.pdf