Around the world, as countries work to rollout renewable energy to meet decarbonisation goals, the role of solar data in ensuring the optimal PV system performance is becoming increasingly evident. From prospecting to operation, solar data can streamline development and maximise efficiency.
Solar Power Portal recently caught up with Clean Power Research’s CEO Jeff Ressler, to find out more about how the use of satellite data has developed, and how it’s helping tackle key challenges from new terrains to balancing the energy mix.
Could you tell me a little about the origin of Clean Power Research and SolarAnywhere?
Clean Power Research was founded by Tom Hoff, after his neighbour asked the question ‘what is the value of solar to me?’
During the 1990s, this was a fairly rare question, because not a lot of people were doing solar. Tom had just completed his Ph.D. at Stanford and with utility experience earlier in his career at Pacific Gas & Electric, he had the background to answer that question. So, in 1998, he embarked on consulting for national labs and utilities in the United States through his newly founded company, Clean Power Research. He had a strong focus on the economics of renewables.
Solar data is a key input to calculating the value of solar, so during this time, Tom connected with Richard Perez, who was—and still is—the leader in modelling irradiance, especially using satellite data. Richard is French by birth and studied in France, before coming to the United States for his PhD. In fact, the model commonly used today for solar sky modelling for irradiance is called the Perez Model, named after him. Their collaboration resulted in SolarAnywhere’s launch in 2008.
We're pleased that Richard is still closely involved with the company. We fund his research lab, and it’s been a prolific relationship in terms of the papers, research and innovation that’s come out, including what we’ve been able to integrate into SolarAnywhere.
Could you tell me about your SolarAnywhere product?
SolarAnywhere is a family of satellite-derived solar data and intelligence products designed to support a variety of different stages across the solar lifecycle. Our customers include global IPPs, asset managers and consultancies developing, operating or transacting on GWs of solar assets around the world.
During the inception phase of a project, we have solar resource data products that support prospecting, as well as engineering and financing a project once a site is identified. Our data helps engineers answer questions like: How much solar can we put in? How much will my design produce at this location, given the specific solar panels and configuration? Then moving into the financing piece, it’s important to show that the predictions are reliable. Can you take them to the bank, show them to the project investors and know they’re going to be trusted?
Once a solar plant is operational, the next phase is benchmarking and forecasting performance. This is where ongoing solar data, including real-time and forecast data, become very important. Forecast data, for example, helps plant operators understand what the plant is likely to do over the next few hours or even days ahead. Often this information is fed into energy trading and resource scheduling decisions made by grid operators.
SolarAnywhere real-time data is used by residential and commercial solar asset managers to compare what the system telemetry is telling them versus what they think the system should be doing. If those measurements are mismatched beyond a certain threshold, decisions can be made about whether or not a service call is required to inspect or repair the system.
What sets SolarAnywhere apart from other data solutions out there?
There are a few key things that set SolarAnywhere apart. One is that we’re a well-established provider. For the past 15 years, we've been very focused on SolarAnywhere’s accuracy and bankability. That means the ability to take the data to a financing entity, demonstrate its provenance, and show that it's been thoroughly validated, reviewed and benchmarked. That has led to high levels of trust by project development firms and their investors.
SolarAnywhere also stands out for its solar data intelligence offerings. Not only are we providing irradiance and weather data, we help our customers with energy modelling so they can more easily and accurately understand nuanced impacts to energy production such as albedo, and snow and soiling losses. We do this using models from an open-source programming library for PV modelling called pvlib. Our customers benefit from the many wonderful advancements coming out of pvlib and the ability to see the algorithms behind these loss and performance estimates. It's not a black box, and that enhances the trust level.
It’s important to help analysts, modelers and engineers be as efficient as possible, so we integrate SolarAnywhere into many of the tools they’re using every day, such as PVsyst, RatedPower’s pvDesign or Valentin Software’s PV*SOL. These are just some of the products that can call directly into SolarAnywhere to access our datasets. All of our products are accessible via an API (application programming interface) as well as direct download from SolarAnywhere. This is really what allows SolarAnywhere to be deeply built into the workflows and operations of our customers.
What do you think the biggest challenges are that solar companies are trying to tackle by using SolarAnywhere?
The state of the art is advanced, and the documentation on how to do a lot of this modelling is fairly well defined and available online. A lot of it comes down to nuanced questions. For instance, if I’m using bifacial panels, what's my albedo? Do I need to worry about things like snow, wildfire or extreme weather events? How does that impact my PV production? Some of our customers are dealing with these types of situations and they come to us and say, 'Hey, we need help with bifacial' or 'we need to be able to model soiling better'. And so we've invested heavily there.
We’re also seeing a shift to firm power, where the solar plant operator takes on more responsibility for weather risks. This is driving changes in the way projects are designed, financed and insured. For instance, we’re seeing growth in solar plus storage plants, which adds a whole new dimension to the design and operation of a plant. The ability to characterize the full-range of possible generation scenarios and forecast accuracy at a given location can play a major role in optimizing plant design and reducing costs. We’ve already introduced features, such as hindcasts, to address these types of issues, and we’re continuing to advance our SolarAnywhere offerings to support the solar plant of the future.
With an increased focus on data, how much do companies need to be concerned about cybersecurity?
Cybersecurity is a concern. We work with a lot of utilities, and if there are vulnerabilities that can make the power going to consumers unreliable or unsafe, that's the worst case scenario. It should be viewed as such by anybody building or operating a solar plant, as well.
We spend a fair amount of time on SOC 2 cybersecurity certification. That involves an annual compliance audit of a wide range of controls that dictate how we manage data, and how systems can be accessed, ranging from physical to online security. These security efforts are a complement to things like GDPR, which we follow for our operations in Europe. Privacy is an aspect of cybersecurity from our perspective, and is part of maintaining the security and integrity of the systems and any customer data that might be there.
What is next for SolarAnywhere?
First we’re growing our SolarAnywhere team. That’s happening in terms of software engineers, researchers and customer-facing staff, including in Europe. As mentioned earlier, we’re focused on helping the solar industry better understand how to deliver firm power. This, I think, is especially important in Europe given the volatility in the energy markets right now and the long-term challenges around a successful energy transition. There is a focus on what these plants actually do, how reliable they can be and when can they provide the power that they're offering.
The ability for solar operators to be able to predict what their systems are going to do is crucial. We’re raising the game by advancing our energy models with pvlib. We’re also focused on delivering high-accuracy, high-frequency data, which becomes more important as solar constitutes an increasingly larger share of the resource mix. This is especially important for applications such as storage plus solar, as it provides greater insight into intra-hour variability, which is needed to optimize the plant design.
We have and will continue to focus on delivering the types of data and energy intelligence to support these new scenarios, as well as validating and updating our validation approaches so that our customers can trust the reliability and accuracy of SolarAnywhere around the world.