Solar LCOE, a Data Mining and Fact Driven Challenge

Fully identifying and understanding real life cycle solar PV costs creates a challenge; while not addressing those costs in detail accurately and dramatically, increases project risk. The probability that one can effectively cut costs without clearly understanding what the underlying costs are, is exceptionally low. This creates a fundamental PV industry-wide challenge. As a result, it brings us to one of the key fiscal performance indicators and benchmarking tools, the levelized cost of energy (LCOE).

However, when it comes to accuracy, where does that information come from?

What is assumed and what is real? … and, if one or more of the major elements is not accurate, what does that mean for the project and its future?

How important is real accurate data versus estimated/assumed data?

LCOE for a project as defined by NREL as used in the System Advisory Model (SAM) software is:

“The LCOE is the total cost of installing and operating a project expressed in dollars per kilowatt-hour of electricity generated by the system over its life. It accounts for: Installation costs, Financing costs, Taxes, Operation and maintenance costs, Salvage value, Incentives, Revenue requirements (for utility financing options only), Quantity of electricity the system generates over its life.”

Wikipedia states that LCOE is seen as:

“The levelized cost of electricity (LCOE) is a measure of a power source which attempts to compare different methods of electricity generation on a consistent basis. It is an economic assessment of the average total cost to build and operate a power-generating asset over its lifetime divided by the total energy output of the asset over that lifetime. The LCOE can also be regarded as the minimum cost at which electricity must be sold in order to break-even over the lifetime of the project.”

Historically, the utility industry has used real measured historic data to determine what the LCOE is for a project. O&M costs are well-documented so when they talk about levelized cost, they are using consistent language and data sources. At this time, this is not the practice in the PV industry because data is scarce.

It makes one ponder. If the LCOE is not based on accurate data, what happens when this cherished metric is based on poor, bad or inaccurate data? And if this is truly the case, how do we correct it?

If indeed the data is good/precise, PV system O&M costs would be precisely projected, under control and accurately reflected in the LCOE equation. The data would be reflected in the numbers used while coming very close to what is observed in the field. Yet today, little industry wide data is either available or shared.

At the same time, the information coming from the field indicates that O&M costs are not under control, while deferred and ignored maintenance are a de facto standard. If the industry continues to wish to effectively drive down the costs without understanding what the costs are in the first place, it would appear that a change needs to take place.

This dilemma is exacerbated by the fact that project financing is highly dependent on those values, which are calculated for LCOE, and therefore dramatically increase the risk to a project when the numbers are not adequately vetted. The results must be relatively accurate metrics and analysis for what project life operations and maintenance costs really are to reduce financial and project risk.

Digging into the data, which is seldom effectively carried out, we see that the primary issue here is the assumption that performance metrics reflect O&M costs as driven by plant health and condition. There is often an assumption that the simplicity of PV plants does not require that robust level of data collection or analysis. These two assumptions are fundamentally flawed and thereby increase project risk.

Part of the dilemma here is if PPA, system owners and asset managers do not believe they need that level of information, they will not invest in the essential or accurate equipment or require the data necessary to accurately assess risk. To make matters more complicated at this time, they aren’t prone to share cost data.

It is clear today that to accurately address risk, additional investment in systemic health and condition sensing metrics and analysis are crucial. This would logically lead us to the proposition that today’s published costs for system delivery, which includes all costs, is in all likelihood fundamentally flawed and inaccurate. Therefore, some level of adjustment might be made at all levels of the industry to understand that there is a reason to collect, curate and share data properly to ascertain all of the real costs for operations and maintenance. It dramatically addresses the need to focus on plant reliability, availability, maintainability, testability and safety.

To achieve cost-effectiveness, further steps must be made that address how data is going to be collected, what data is essential and how it is going to be analyzed. Buying the least expensive monitoring system does not achieve that goal. Beyond that, how will results be effectively communicated to all of the appropriate stakeholders? And how are they going to use that information?

One is drawn to the conclusion that today’s methodologies must be further developed and enhanced. This can be achieved with advanced technology, systemic sensing and broader EM spectral analysis delivered by the addition of advanced mobile and aerial robotics to do much of the most critical data collection. Nevertheless, this still requires a far greater level of analysis and automation.

Yet, we’re still stuck with the assumptions made by many industry leaders where we have not come to terms with the basics of O&M on project lifecycle costs. We have not come to terms with the fact that many O&M costs can be designed out of a plant with more robust specifications prior to design combined with an enhanced quality assurance commissioning (QAC) prior to project turnover to the owners.

This fully vetted awareness of costs is “The Standard” in most other successful technology industries while their acceptance within the PV cosmos is well overdue. So maybe, before we jump to the conclusion that “it costs too much,” we might ascertain what the true costs are as balanced against the real savings. That’s just good business, the cornerstone of both technological maturity and profitability. The fact is these steps are a matter of when these measures are included, not if.

The reality is that those organizations that are going to own the projects they build will be the first to adopt these new technologies and standards. Remember, the standard that exists in the rest of the technology world is a reality that has allowed for profitable and sustainable growth. Their success and info is shared.

Organizations that continue to develop small, medium and large PV projects based primarily on the least cost model will be driven out of the industry as not being cost-effective. This is nothing new!

What the future holds is the adoption, not of new processes and procedures, as much as the adoption of historic and existing processes and procedures that worked for the utility and many other technology industries. That is a tried-and-true approach that is well worth understanding, considering and applying.

Previous articleNY Green Bank’s Clean Tech Investments Reach $346 Million
Next articleNYSERDA to Fund Emerging Energy Storage Technology
John R. Balfour, MEP, PhD, is President and CTO of AstroPower Corp. Dr. Balfour has spent 32 of his 40 years of PV experience as an EPC and has been a PV energy consultant and author since 1977. AstroPower is a consortium of specialized, experienced PV- and technology-related organizations dedicated to making PV projects more efficient and profitable throughout their lifecycle. Contact:

No posts to display