Albuquerque, NM | 5/1/16 –
Thomas Conroy, President, Array Technologies, Inc.
Solar power plants have taken their place alongside coal, natural gas, nuclear, and wind as legitimate and valuable generation resources. Solar comprised 29.4% of new electric generating capacity in the U.S. in 2015, exceeding the total for natural gas for the first time. 1 The acceptance of solar by utilities has happened seemingly overnight, moving from an annual utility-scale installed capacity of 1.8 GW in 2012 to 4.2 GW in 2015. Solar is following wind power by a few years in its meteoric path to volume, and has the opportunity to avoid the difficult early stage lessons that wind and nuclear industries experienced the hard way. The methods and roadmap to reduce long term power plant risk is clear: test new technologies before deploying in volume, establish consistent industry standards for structural design, and utilize well established technical predictive tools to evaluate and account for long term maintenance costs. Failed companies and impaired project assets which act as a brake on industry growth can be easily prevented.
$388.57 billion was invested in utility-scale solar from 2011-2015 to finance, construct, and grid connect power plants around the world. 2 Owners will soon start focusing on optimizing those assets for both operational and economic performance. The product architecture and technology choices that were made up-front will determine the long-term equity owner’s returns, debt holder’s coverage ratios, and utilities’ PPA contractual fulfillment. There are many important stakeholders, so getting the up-front decisions right matters.
In the rush to continue solar’s impressive cost reduction curve, naturally a great deal of attention is focused on initial installed costs. While initial costs are unquestionably important, a more useful evaluation metric for the 30-year lifetime power plant business is the Net Present Value of the project, equivalently expressed as the oft-used levelized cost of energy or LCOE metric. Since 20 to 30-year renewable PPA’s carry minimal inflation protection, or are even fixed over the lifetime, it is crucial for equity owners that the 30-year operational costs are accurately estimated and hold no surprises. Project values quickly deteriorate if the expected operations and maintenance costs rise even a small amount from what was originally projected, with the equity holders taking the first hits to their expected returns.
The risk of dealing with unexpected costs is not a hypothetical situation that hasn’t happened before in renewable energy. From 2006 to 2008, during wind generation’s initial boom, large utility-scale turbines from new market entrants were leaping from design to volume deployment overnight, with no field experience possible given the rush to deploy. After these immature designs were deployed in the field, a long series of design and manufacturing issues revealed themselves which caused upwards of a billion dollars of field retrofits and extensive downtime. Although the turbine OEM’s such as Suzlon, Gamesa, and Clipper did pay hundreds of millions in field retrofit and liquidated damages (LD’s) for lost production, the LD’s were typically not enough to make the project owners financially whole and long term reliability was questioned. The ultimate outcome was reduced project values, while the turbines which were found to be of questionable quality and poor reliability became difficult to finance in the U.S. market.
The technical methods and tools industries use evaluate and ensure quality and reliability levels are well known. For consistent risk mitigation for structures, other industries require design certifications to one of several global standards bodies, which will define dozens or hundreds of structure specific static and dynamic load cases which must be rigorously met. These load cases will typically define extreme wind and snow events, including combined wind velocity and directional changes, laminar/turbulent flows, changes due to ground feature interactions, and other conditions that are likely to be encountered over the 30+ year life of a power plant. For prediction of long term reliability performance, there are multiple tools from simplistic probability to highly complex mil spec standards used to estimate the long term reliability performance of complex systems. These tools will focus on component counts and their associated reliability, as well as the impacts of multiple cascading failures if those cascading failures impact the risk to structures or life.
Some of solar generations’ cost competitiveness has been gained by increasing scale. Module rows, currently at 80 meters and going to 90 are now approximately 50% longer than the wingspan of a Boeing 787. Increased wind and snow loadings have become critical factors to consider, as is the question of whether the structure has been designed to intrinsically handle the design loads or relies on back-up systems needed to move it to a “safer” position to withstand weather events. The assumption that all high wind events at solar power plants occur with laminar horizontal flows needs to be supported at the increasingly varied solar plant sites across the country.
Long term reliability tools will reveal systematically what is common sense to many; which is that the most effective way to dramatically increase reliability is to reduce parts counts. Across different tracker architectures, the parts subject to failure can range from 187 to more than 29,000 in a 100 MW plant. Simple predictive failure and economic tools tell us that a parts disparity of this magnitude is likely to result in more than $0.03 / W (dc) difference expressed in present value terms. This can amount to unexpected annual maintenance costs of more than 3% of revenues just for the tracker. These inescapable parts-count related failures and expenses lead successful companies to pursue parts count reduction as a corporate strategy. We’ve seen this strategy play out in the disk drive industry, where product failure rates have decreased from 30% per year in the early 1990’s to today’s performance of less than 1% per year failure rates, principally driven by parts count reductions. Another example is Subaru which pursues parts count reduction as an explicit corporate strategy to increase the reliability of their products. The principle is simple: if there are fewer parts there will be fewer parts to fail.
One of the most extraordinary and unique aspects of solar generation technology is that it generates electricity with little to no moving parts. Suggesting to a utility that a solar power plant could be operated for 30 years with no maintenance would cause your audience today to chuckle and threaten your credibility, but that is exactly the opportunity that systems with minimal failure-prone parts ultimately can offer. Readers of this article will recall that five years ago the idea that utilities would be building cost-effective solar power plants on a massive basis would also have been laughably received. Of the three key components of modern solar power plants – modules, inverters, and trackers – only trackers have a significant number of moving parts, principally because that is their function. With advances that have been made in utility tracker technologies by industry leaders over the past thirteen years, it is now possible to provide tracking systems with zero scheduled maintenance over the expected 30 year project life. To achieve this remarkable outcome requires adhering to disciplined architectural and design principals up front, such as designing with inherent structural strength, minimizing parts counts, and choosing technologies with field proven long-term reliability.
The outlook for solar generation could not be brighter. The industry has now brought in many new participants and achieved an astonishing volume ramp driven by cost reductions. Many of the industries new entrants bring fresh perspectives and approaches which are welcome and invigorating. The ultimate winners who will drive the industry to the next level, however, will be those who are focused on 30-year economic value and how to optimize both up-front and operational costs over that time span. Solar technologies offer the opportunity to reach unprecedented lower maintenance costs and risk profiles than any other generation technology. The solar industry and its participating companies should now begin to utilize the readily available methods and tools to reach those levels of lifetime efficiency that will add one more significant advantage to solar generation. Doing so will not only help the industry avoid the setbacks that other generation technologies have experienced following their breakneck growth phases, but it will further solidify solar generation as the inarguable choice for new daytime generation.
- GTM Research, U.S. Solar Market Insight, 2015 Year in Review
- BNEF, Clean Energy Investment, By the numbers – End of year 2015