Keeping an eye on performance

Utility, consultant and technology manufacturer representatives echoed a common theme at the 4th PV Power Plants Conference, which was hosted by Solarpraxis on November 28 and 29, confirming that more accurate data gathering and analysis is needed to help augment utility-scale PV plant operation and maintenance, as well as to optimize the value of assets over the long term. Speakers paid recognition to the fact that the U.S. PV industry is still in its infancy, and that much work remains before plant material configurations are more standardized. Such coordination will be critical before performance data is widely shared and industry best practices are perfected, they lamented.
What utility-scale PV system owners are aiming to avoid by improving data acquisition and management can be likened to “death by 1,000 cuts,” equates Hugh Kuhn, a managing member of HAK Solar Advisory Services of Mill Valley, California. These performance cuts can translate down to individual insect invasions, small blobs of soilage or nascent rust points, he pointed out. Cumulatively, the diverse universe of such drags on performance threaten to limit production and profitability, percentage by percentage, he noted.
Among utility representatives, Josh Gerber, the principal architect of San Diego Gas & Electric’s smart grid, explained that the variations in weather across the vast territory served by his utility – ranging from the Pacific Ocean to the desert, and from mountains in the north to the Mexican border in the south – has literally created hundreds of separate sub-climates to contend with, served by a myriad of generation asset types, including several varieties of solar generation. Overall, the utility interfaces with over 20,000 distributed generation systems, most of which are photovoltaic, he said.The mathematical permutation of such a combination of module, inverter and other equipment components, multiplied by broadly different weather conditions spells out the need for an industry-wide database of performance data, specific to equipment and location, pointed out Benjamin Compton, the Chief Operating Officer of Meteocontrol North America, based in Alameda, California. His company has been building a database encompassing many such variations, and the use of such benchmarking can help plant operators and planners increase operational and financial yield, he suggested. While the U.S. Department of Energy’s National Renewable Energy Laboratory has been working on the same sort of database, he noted, their formidable effort has already been a decade in development without a final product in sight.

Performance data need

Another provider of performance data likened the need for information in the PV industry to that of the automotive industry, especially when it comes to assessing the value of aged assets: “We act as a third-party witness to performance data; it’s like having Carfax, which tells you when the oil was changed and what repairs were made on a used car,” said Robert Schaefer, the CEO of AlsoEnergy, based in Boulder, Colorado. His company provides energy monitoring and financial management software applicable across the PV supply chain.
Selecting a PV performance database to trust can be a vexing question in the U.S. market thus far. “The whole PV value chain needs data, and against what benchmark you compare your data is a very challenging question in the industry today,” surmised Casey Miller, a cofounder of Envect, a renewable energy consultancy based in Fort Collins, Colorado.
One technology manufacturer touting module-specific data was Paul Grana, Director of Marketing at Tigo Energy, in Los Gatos, California. He noted that his company’s in-module electronic monitoring board is enabling operators to utilize a module-specific performance profile in place of generic IV-curve tracers during commissioning and throughout operation to better track variations. Through Tigo’s collaboration with Folsom Labs, of San Francisco, PV system owners can also predict the precise length of time required for return on investment for individual modules, and for a combined system, he noted.
Another monitoring system provider, Murad Can, the Vice President of global customer operations at Skytron Energy of Berlin, reckoned that PV system owners who fail to monitor performance typically lose three to six percent of their potential performance efficiency. The cost of monitoring, he suggested, should be compared with relative loss-avoidance: the cost of the detection of a string failure can pay for a monitoring system in two years, he estimated.

Tying warranties to performance

Other companies that provide monitoring services, like Swinerton Inc., of San Diego, a broad engineering, procurement, construction (EPC) service provider, trust in themselves more so than in others. Joe Brotherton, the Manager of Technical Services for Swinerton, in San Diego, indicated that monitoring and response capabilities are so important to his company that they have built their own proprietary monitoring and analysis system, which is not for sale to outside users. “Contracts these days say that the EPC must guarantee a certain performancelevel for a PV plant over a certain period of time, so we’ve chosen to control everything we can, because our own bankability is on the line,” he said.
Performance data may become more important in the U.S. utility-scale PV market sooner rather than later. Several consultants speaking at the conference pointed to the rising importance of performance data relative to the maturation of U.S. utility scale plants, in which investors are required to hold their position only five years to fully recoup the federal tax credit provided at construction.

Using data to value assets

Several speakers urged potential used-plant investors to trust in data. Laks Sampath, Executive Vice President of Neozyte of Palo Alto, for example, stressed that without detailed performance data, a prospective buyer of a five-plus-year-old PV plant may need to perform a basic recommissioning of the plant to determine its true value at a proposed purchase date. “The culture to ensure long term project success is not there in the industry yet. Too many players just want to make their money and get out,” he admonished.
Utilizing performance data to optimize operations and performance also clearly depends on economic projections, which can take into consideration a wide range of factors. However, “the most important factor in performance ratio planning is weather data,” suggested Rusty Sage, a senior operations engineer at Gestamp Solar, in San Francisco. His company is a developer and operator of PV plants, and participated in the development of 400 MW of PV last year. He also pointed out that insurance protection against weather can be crucial for an operator in the case of lightning damage, for which loss of business coverage must kick in within a matter of days – not several weeks – to prevent catastrophic economic impact.

Validating predicted data

Unfortunately, baseline data on PV equipment failure is still scarce in the U.S. industry, observed various speakers. One materials solution provider representative, John Reimers, a downstream marketer at DuPont Photovoltaic Solutions, of Wilmington, Delaware, noted that his company has scoured the world to develop a database of some 300 failed PV modules from different climates and different geographic locations over the past two years “to help the industry understand what is going on in the field,” he said. “We are seeing massive changes in the way modules are being built, as the race goes on to cut module production costs,” he noted. “As of September of this year, there were still 61 module makers in the United States, making 900 different kinds of modules,” he asserted. “What the industry needs is a production material specification sheet for every module, to standardize knowledge about what is being bought,” he suggested.
Europe may ultimately provide the lead for the U.S. database building effort. “We are beginning to look at operating data from Europe, where PV plants have been online longer, and the information is helpful,” noted Richard Dovere, the Managing Partner of New York-based Adamas Energy Investments, an investor in distributed power generation assets, including PV.

Early data management adoption

By including better performance data in the planning stage of a utility-scale plant, performance may be optimized just enough to enable a developer “to lower the price bid on a PPA by a penny per kWh,” which could prove to be enough to win a competitive deal, pointed out Lee Feliciano, a project developer at Kyocera, based in Phoenix. “As the tax credit is phased down to 10% in 2016, encompassing performance data will become even more critical,” he predicted.
This belief was reiterated by Compton, of Meteocontrol: “The use of technical due diligence during the project planning stage prior to investment can increase plant operating performance yield by four to five percent, by way of continuous quality assurance,” he suggested. “We look at 500 different criteria to analyze predicted and actual performance in a plant, and provide a technical rating based on the same scale as the international financial ratings companies,” he pointed out. “We have secured accreditation for our criteria assessment from DAkkS, the German federal accreditation agency,” he noted.
The standardization of project design and the performance data associated with it will greatly facilitate financing in the future, suggested Craig Robb, the Managing Director of Zions Energy Link, a unit of Zions Bancorporation, in Phoenix. “I am bent on standardization, which really doesn’t exist in the utility-scale PV market yet, so the answer to the problem is repeating a plant model that customers are already comfortable with, which reduces the financial transaction cost,” he explained.