Sunday, December 13, 2009

Design for Solar Investment: Fitting the Inverter

Supersize me!

The engineers that deliver energy solutions to homeowners and small businesses labor under demands much different from those experienced by the engineering teams that design large, commercial solar energy systems. For smaller systems, time, money, and effort all become more personal and immediate; if not more precious.

A solar electric energy system is an investment that can yield extraordinary returns in sunny climes. Even where the sun is less ample, energy generated from the sun can have extraordinary value. While a small solar investment may not require funds on the scale of a 401K, it is still an investment that requires significant capital even where rebate incentives apply. Regardless of the size of your solar investment, the system must have a cost-effective design in order to attain the highest possible return.

A cost-effective design will be the least expensive design that provides reliability and durability while maximizing lifetime energy production. In operation, it is important to ensure that each and every watt is pulled from the system throughout its long life in order to avoid the unnecessary purchase of electricity from the grid. Energy production is especially important at times when the cost of energy from the grid is the highest; both at any moment of any day, and in later years as energy costs continue to rise.

Even though this all seems straightforward, many designs falter with the fit of the inverter(s). Too often, the selection of an inverter is determined only by its cost and by guidelines that consider power capacity ranges, but that neglect ‘comfort’ and value. The cost and size alone of a pair of shoes will not inform you whether they will carry you happily down the long road of years. To experience real value and to reach your distant destination unblistered, you need shoes with a proper fit.

One design philosophy that fails the rigorous test of value is the practice of incorporating an oversized inverter.

“The Inverter (the heart of the system) is always oversized. By operating the inverter below its maximum power rating, it stays cooler and extends its operating life.” [emphasis mine]
- company website

In one particular instance, following this dogma led to a system design that incorporated an oversized inverter with a retail cost of $1,200 more than the standard inverter. This additional cost represents a 47 percent increase over the cost of the standard inverter, a 6 percent increase in materials cost, and a 4 percent increase in the estimated cost of the installed system. A design that front loads system costs ignores a central principal of investment: the time value of money.

Modern inverters are designed to endure harsh environments. If the instructions for properly mounting and operating the inverter are followed, a mean time to first failure (MTTFF) of ten years is typical. Quality inverter manufacturers now provide ten-year factory warranties. The argument that cooler operation may extend the operating life of an inverter is not without merit. However, even if the site does not provide a shaded location at which to mount the inverter on a north-facing wall near the electrical service entrance, an adequate, inexpensive cooling solution is available for only about $100. A specialized fan costs dramatically less than the $1,200 additional cost for an oversized inverter.

While an approach that focuses on inverter size is faulty, it serves to illuminate a number of important, yet frequently disregarded considerations for fit. A design with an oversized inverter will provide no additional power. In fact, it may provide significantly less.

Transition Power

The inverter's lowest operational voltage (LOV) (a.k.a. startup voltage) is the input voltage at which an inverter starts and stops delivering output power and, therefore, the duration of its operational window. The match of the LOV and the output voltage of a string of (series connected) modules is a critical factor in both maximizing system power output and in determining if any value is achieved from incorporating an oversized inverter into the system.

Even though there are broad ranges of inverters with varying capacities but with similar LOV, the LOV broadly correlates with the size of the inverter. Therefore, a smaller inverter may respond sooner during transitions from low to high irradiance, and postpone a shut down during the opposite transitions.

One might assume that the power generated during these transition times is negligible, and that its value is insignificant. However, the total of transition time is determined by more than sunrise and sunset; it includes many periods of transition between shadow and sunshine. Therefore, potential energy loss during transitions is compounded. It may also be that the energy lost during these times has extraordinary value because of the structure of a rate schedule, consumption patterns, and/or climate. The value of the power generated during these transition periods accumulated over the many decades of the system’s operation can be significant.

If the primary goal is to maximize the return on investment in a solar electric energy system, then a smaller inverter with a lower cost and a lower LOV may be the better option: even if it means using a few less modules. There are also several technological constraints that favor a smaller inverter with a lower LOV.

Temperature Effects

The most obvious constraint is the effect of temperature on module performance. Photovoltaic modules are dynamic, electrical devices with imperfect efficiency. Therefore, they dissipate energy in the form of heat. Because of albedo and to a much lesser extent electrothermal generation, photovoltaic modules operate at temperatures above the ambient.

The practice of incorporating an oversized inverter increases the risk of the voltage of a string or array falling below a larger inverter’s higher LOV. This risk is exacerbated by several factors. First, a module’s operational voltage is lower than its nominal voltage. Second, a module’s maximum power voltage (Vmp) will be lower than its open circuit (Voc) voltage. Thirdly, the voltage of a module operating under high temperature is lower still. In the desert and other regions with hot climates, when outdoor temperatures are extremely high, electricity (for cooling interiors) is particularly valuable. A power loss due to a clipped power generation window during the summer is especially expensive.

When summer ambient temperatures regularly approach 50° C in the desert, then module operating temperatures can become very high. Even though morning temperatures at the onset of the power generation window may not rise this high, power transition periods can occur at any time due to weather, shading, or other factors. Furthermore, when overnight lows are above 35° C (95° F) it does not take long for module temperatures to rise in the morning.

Module Degradation

Another constraint is module degradation. The performance of silicon solar cells and other photovoltaic materials begins degrading naturally after manufacture. The degradation of materials, cells and assembled modules is accelerated when a solar electric system is installed, exposed to the elements, and energized.

The output current of each module in a string, array, or block will diminish as degradation proceeds. In contrast, the voltage (Voc) of most modules in a system will remain relatively constant as they degrade over the years. Still, over the system’s long life, module degradation increases the risk that the voltage of a string or an array will increasingly sink below the LOV of the inverter in a poorly-designed system.

The importance of degradation in system design is amplified because it is highly likely that the cost of energy from the grid – the costs that are expected to be avoided by replacement with solar energy - will increase over time.

Module Mismatch

Yet another important constraint is module mismatch. Any particular model of photovoltaic module is manufactured to tolerances that are typically specified as a percentage range that is evenly distributed around the model’s nominal peak power at Standard Test Conditions (STC). Often, this percentage tolerance is the same for models of dramatically different sizes. A 225-watt module with a ±5% tolerance will cover a range of 22.5 watts and a 95-watt module with the same tolerance will cover a range of 9.5 watts.

For many manufacturers, improvements in quality control have led to improved tolerances. A ±5% tolerance is commonly specified. Yet, a drop of 5% below the nominal peak power in the performance of modules in a string or an array can have significant negative effect on the performance of a system that are magnified by other detrimental factors.

Environmental Factors

There are also a bevy of environmental factors that can cause system performance to deteriorate in both apparent and subtle ways. Dirt, cable and contact wear, moisture incursion, tree growth, and critter nibbling can all exacerbate the technological constraints.

Risk Engineering

Even with the potential of any particular constraint to damage the performance of the system, the greatest conceptual hurdle is not the technology or the physics, but the probabilities. The probabilities of any constraint occurring must be combined with the probabilities of the magnitude of the occurrences, the probabilities of the speed of degradation, the probabilities of the constraints occurring in combination, and the probabilities of the magnifying effects of their interactions.

Anomalous modules may appear. Occasionally, a module may experience an extreme degradation. Or, an anomalous module’s degradation may express itself as a voltage drop that is apparent only during periods of power transition or under extremely high temperatures. A module may be delivered within stated tolerance, and even remain within tolerance for many years before it exposes a flaw. The probability of anomalies appearing increases with temperature and time.

Even though the risk of an anomalous module appearing increases with the number of modules, the effect of this risk on the duration of the power generation window is counteracted by the greater string voltage. This is also true of module mismatches. Therefore, the overall risk that module degradation will affect the power generation window is higher for smaller, lower voltage strings. For example, choosing to use eight modules instead of ten modules in order to save a few dollars of capital expenditure may result in a more rapid approach to the LOV and an inordinate diminishment of the solar investment.

Another risk to the solar investment is the eventual repair or replacement cost if the inverter fails and the warranty has expired. Will the costs of repairing an oversized inverter be higher than repairing a smaller inverter? Years later, will an unsuspecting owner replace a failed inverter with a similar model, not realizing that the larger model is now too big for the degraded array? Will future technology solve this potential problem?

The Timing of Generation

Further complicating matters is the economic issue of the timing of generation in relation to consumption. For some, early morning and/or late afternoon generation is more important. In the absence of solar trackers, module technologies whose energy production is less dependent on angle-of-incidence have an advantage, but only if the inverter can respond. There may also be a seasonal consideration where increased cloudiness may inordinately diminish production because of a pinched power generation window.

Building an Industry

A cost-effective solution must consider all these interacting factors. System designers sometimes use open circuit voltage (Voc) as a design parameter without considering the ramifications of maximum power point tracking (MPPT), module temperature, mismatch, degradation, and environmental factors for string voltage over the life of the system. Even before degradation occurs, there can be more than a 20 percent voltage difference between Voc at standard test condition (STC) temperature of 25° C (77° F) and Vmp at module temperatures of 50° C.

When designing a system it is important to be both knowledgeable and flexible. Every situation presents unique challenges that may require unique solutions. It may be the case that future expansion is a part of the design strategy. In this case, an oversized inverter might be a valid solution. But, “always” is a very risky design strategy. A cautious “always” is better applied to product, deployment, and operational strategies. Incorporate proven components in systems. Modules should have IEC[1] performance qualification in addition to the required safety certification(s). Design systems for quick, repeatedly dependable installation. It is (almost) always a good strategy to monitor peak power and power generation —with special attention to power transition periods — throughout a system’s life. Finally, provide responsive service should the system’s performance falter. Responsible companies design and build not just superior solutions for their customers, but also work to build the reputation of our very important solar industry.

[1] International Electrotechnical Commission.


  1. Wow, this is a great overview of the all the issues one needs to take into account when digging deep into the details of PV system. Thanks!

  2. You're welcome, Lon, and thank you!

  3. Solar energy is a great idea and it is also an environment-friendly.

  4. It is rare that risk management is discussed in relation to small energy system design. I would like to read more. Can you recommend a source?

  5. Like you said, voltage from the Solar Inverters being too high, could decrease efficiency, how much? Voltage range is 250 to 600 volts, and temperature range is -20 to 45 Celsius. What is a optimal output?


  6. milan,

    The article makes the point that there are many risk factors to optimal performance. What is considered 'optimal' during commissioning may not be optimal a decade later. Design should consider performance throughout the system's life-cycle.