Basic Approach and Methodology for Photovoltaic Module-Level Techno-Economic Analysis (Text Version)

This is the text version for a video—Basic Approach and Methodology for Photovoltaic (PV) Module-Level Techno-Economic Analysis (TEA)—about analyzing model reference designs, manufacturing processes, and PV modules costs.

It’s Part 1 of NREL's Solar TEA Tutorials video series.

[Audio begins]

In this section, I will review our procedures and some suggestions for completing techno-economic analysis, or TEA, at the PV module level. So, in this section, the focus will be a little different than TEA for the PV system level, but that will be covered in a later section.

What Is Techno-Economic Analysis?

So briefly, what is techno-economic analysis? If you break down the name, it refers to the combined analysis of both technical and economic factors. So, for example consider the technical and economic effects of thinning a solar cell. You have a thinner solar cell, so it's using less material, which means your material costs per cell are lower. However, what are the effects of thinning on the efficiency of the cell? If the cell is no longer optically thick, efficiency will decrease, so your cost per watt will go up. TEA accounts for both the effects on the technical performance and the economic performance of thinning a solar cell, so it can quantify the trade-offs and identify an optimum thickness.

The Analysis Process

So, here is an overview of the content I'm going to cover in this section of the tutorial.


I'm just going to jump right in with scoping. In order to complete the subsequent steps in the analysis process, you have to scope your TEA first.

[Slide lists the following questions: what questions am I trying to answer, who is the primary audience, what stage is my technology or idea at?, what resources do I have (time, money) for the analysis?]

To scope your cost model, you should identify the questions you're trying to answer. Maybe you're interested in identifying the highest cost categories of your technology in order to highlight areas for improvement, or you want to map the anticipated trajectory of technology progress, or compare the techno-economic performance of different materials or processes. You also need to identify the primary audience of your TEA results. For example, researchers will often have different areas of focus than manufacturers. You should also consider the current stage of your technology. What scale is your technology at now, and what scale are you interested in? Laboratory unit costs and performance are typically going to be higher than the costs and performance of a commercialized product, so modeling at laboratory-scale will appear more expensive. If laboratory development of your technology is still in process, you may want to model theoretical potential. However, that will require significant assumptions. Finally, you should consider the resources you have at your disposal, such as time and funding. This can limit the degree of detail in your model, or how many iterations or scenarios you perform. You may have to prioritize some areas for high detail and rely on assumptions for others.

So now I'll review considerations when selecting reference design scenarios and iterations and process flows for your cost model.

Choosing Reference Designs

The first step is to choose one or more reference designs. Depending on your scope, you may only need one. Some options for selecting a reference design include performing a survey of products in the market and defining a design that is generally reflective of the market; or you can consult industry members or other experts that can describe or identify designs that are appropriate or generally reflective of the market. You could also conduct a literature review. However, be aware that some device designs or aspects of devices in academic literature are not always reflective of what will be used if that technology were produced commercially. If you are planning to publish your TEA results, it is best to select reference designs that are not specific to one company. This helps protect proprietary information, avoids the appearance of marketing a particular company, and will give your publication broader impact.

Selecting Scenarios

After selecting reference designs, you can then define any necessary scenarios for your techno-economic analysis. Different scenarios can include modeling production at different volumes to show economies of scale or map a trajectory for the technology. Modeling different manufacturing locations will reflect local differences in costs such as labor and electricity rates. You could also vary design parameters such as product size or material compositions, or process options such as deposition methods or automation.

Developing Process Flows to Use for Analysis

Finally, there are other aspects of process flows that must be selected which vary significantly at different scales and stages of production, such as the through-put of production equipment, uptime, or yields.

Creating Cost Models

Now I'll review the building process and financial structure of the cost models.

NREL’s Methodology and Alignment with GAAP and IFRS

[Slide shows diagram that consists of the following: cell and module technologies, step-by-step cost of ownership inputs, and GAAP and IFRS standards, all of which equal total module supply chain costs.]

I'm going to walk through a diagram of how different parameters get filtered through our cost model. First, I'm showing a list of many different module and cell technologies that our team has analyzed. For each of the technologies listed on the left, you have to collect the data listed in red in the center, referred to as cost of ownership data. As we've explained earlier in this section, this data will vary significantly based on what scenarios you are considering. Finally, the cost of ownership data is structured into a cost model using the techniques on the right, including generally accepted accounting principles and international financial reporting standards. These standards classify costs into three main categories. Variable costs are tied to output, while fixed costs remain constant regardless of output. Together, variable and fixed costs comprise the cost of goods sold, or COGS. Additional expenses not in COGS include research and development expenses as well as sales, general, and administrative expenses, or S,G, & A. These three categories estimate total module costs. It is important to note, however, that the concept of module costs, or total module costs, is not the same as the concept of module price, which we will cover in a couple slides from now.

Quick Intro to Basic Finance Terms

So, we will show a short video of how to go from total module costs to a modeled module price, but first I'm going to go over a few basic finance terms that you might need to be familiar with when working with a cost model or looking at TEA results. I don't think anything in this tutorial will be dependent on your understanding of these terms, but you may see them referenced a few times. So, the first term is depreciation. This refers to the category of fixed costs, such as equipment or buildings, that can be expensed over multiple years. The expense schedule, or depreciation schedule, is set by the IRS for tax purposes. You may also hear the term cash flow, or cash flow analysis, which describes the accounting of revenues and costs over some project lifetime. So, the key point there is that it's over a specific life; it's a dynamic analysis. The discount rate represents the time value of money and is used in a cash flow analysis to discount future revenues and costs compared to the current value of money. So, if you're not familiar with the concept of a discount rate, you can think of it as a similar concept to inflation, where the same denomination of money was worth more in the past than it is today. And then a term called the weighted average cost of capital, or WAC, is often used as the discount rate. WAC is calculated using a company's ratio of debt to equity financing, as well as the respective cost of debt and cost of equity, where the cost of debt is essentially the interest rate on the debt, and the cost of equity is the return rate for investors that hold equity in the company. Okay, so now we'll show the video on how to estimate the product price using our TEA methods.

Minimum Sustainable Price

[Video played, 0:08:50 to 0:13:10]

[Music playing]

[Video narration begins]

Let's say you're working on the next big thing in solar energy. Maybe you're a researcher, wanting to know how your technology fits into the current market. Or maybe you're a manufacturer, trying to identify trying to lower costs. For scenarios like these, minimum sustainable price, or MSP, is a useful metric to consider. Minimum sustainable price is exactly what it sounds like. It's a price that provides the minimum rate of return necessary in a given industry to support a sustainable business over a long term. Specifically, MSP is influenced by manufacturing costs, overhead costs, and other financial considerations such as financing, discount rates, and tax incentives. Let's take a closer look at manufacturing costs, also known as the cost of goods sold. Understanding manufacturing costs can help identify the major cost drivers for a particular technology. Manufacturing costs include materials, labor, electricity, maintenance, equipment costs, and facilities. The next set of costs that influence MSP are overhead costs. These include research and development costs, as well as sales, general, and administrative costs. Overhead costs can vary significantly between companies, as well as over time within a given company. After summing up manufacturing and overhead costs, we then obtain the minimum sustainable price by assuming an operating margin, typically desired when pricing products within a given industry. An operating margin accounts for interest payments, profit, and the corporate tax rate. A sustainable operating margin can be estimated by interviewing industry members or by calculating the price needed for a business to break even over an assumed business lifespan, while adjusting for inflation and the cost of capital. Given all these factors, the MSP is the price you would have to charge to break even at any given point in time. Let's look at a few different ways that MSP can be used. First, minimum sustainable price allows us to directly compare costs of different technologies. Market prices aren't ideal for such a comparison because manufacturers might be selling well above or below their actual costs. MSP focuses on the actual costs of the two technologies, removing market factors like supply and demand fluctuations from the comparison. While you could compare technologies based solely on manufacturing costs, the MSP also takes into account financial factors, such as different financing fees, that can create cost differences between technologies. Furthermore, considering MSP can provide a way to estimate what prices and margins might be for manufacturers when no public information is available.

Another important aspect of MSP is that it will adjust over time, as costs change. MSP is not the absolute minimum sustainable price that could be achieved by a given technology, just the minimum at that time and location. Here you can see how for one technology, NREL's estimated costs, MSP, and market prices have changed over time. Our costs and MSPs are benchmarks, suggesting a typical case within the industry. Sometimes the market price is below the MSP, which can reflect low margins in the industry. A company may also price below our benchmark MSP if they have achieved lower overall costs than the industry as a whole. Additionally, MSP is a good metric for setting policy and cost targets for a technology. Ideally, these targets should reflect actual technology costs, not prices that have been inflated or depressed by market factors. In a nutshell, minimum sustainable price can be a useful metric when comparing existing energy technologies or introducing new ones. To learn more about MSP, see NREL's solar cost analysis applications.

[Video narration ends] 

Cost Model Results Summed for PERC Wafer, Cell, and Module

Here is an example of cost model results, shown here for a vertically integrated manufacturer that produces monocrystalline wafers, PERC cells, and modules. The materials, labor, electricity, and maintenance costs represent the variable cost component of COGS, and the depreciation category represents the fixed cost component of COGS, where that's buildings and equipment depreciation. Then there is the gross margin, which includes R&D expenses, S, G, & A expenses, and the operating margin. We typically estimate R&D and S, G, & A expenses from quarterly financial statements of relevant publicly traded companies and apply them as a percentage of COGS. The operating margin includes interest payments on debt, equity payments for investors, and any additional profit, as well as federal, state, and local corporate taxes on that profit.

Historical PV Manufacturer’s Margins

So, just as an illustration, even though we assume sustainable margins for our cost models, this isn't always the case in the actual market. As you can see here, operating margins are often negative in PV industry. Negative margins can often reflect external market factors such as supply and demand.

Collecting Data

Now I'll just go over some advice and suggestions for how to collect the cost of ownership data that we listed previously. Here I'll review some typical sources for different types of cost of ownership data. If you're looking for equipment costs, your best bet is vendors of the equipment. Many vendors have a cost of ownership fact sheet prepared for prospective buyers. For material costs, there are a few different options. You can request quotes from material suppliers. Make sure to specify what purchase scale you're interested in. You can sometimes find commodity pricing online or through certain subscriptions. Market reports can also provide material costs, but the reports can be expensive, and they don't always have great a degree of data transparency. Finally, for many of the process flow parameters, these are best sourced from manufacturers of the product you are modeling or manufacturers of similar products.

Preliminary Analysis

Finally, once your model is built and data gathered, there is an iterative process to refine and discuss your TEA results.

Feedback Gathering

This includes, once you've generated your preliminary analysis, you then share those results with relevant companies and industry members to gather feedback and critiques.

Iterating your TEA results using feedback from industry can be useful to focus on and discuss areas of more interest, or even catch inaccurate assumptions which will make the final analysis more relevant and robust.

[Audio ends]

To continue with Part 2 of the Solar TEA Tutorials, see Cost Modeling for Specific PV Technologies (Text Version).