Solar-electric systems are becoming more common on homes, businesses, and as large generating facilities. This is partially due to increased incentives and rebates available, which help reduce system costs. However, while it may not be an apples-to-apples comparison, the fact remains that PV-generated electricity can be 1.5 to 4 times the cost of the typical, fossil-fuel-dominated utility blend.
The price per kilowatt-hour (kWh) for a PV system is calculated by dividing the system cost by its estimated lifetime energy production. One way to reduce the cost per kWh is to maximize the energy the system produces. Of course, the benefits of PV-generated electricity go far beyond the system’s initial cost: Each clean kWh produced by a PV system can displace a kWh that would be produced by conventional (polluting) energy sources. So wringing as much production as possible from your PV system makes good sense financially and environmentally.
PV modules typically have warranties of 20 years or more, and expected operational lives of more than 30 years. Over this time, a small amount of lost energy each day can add up to significant losses. Poor installation practices can result in increased maintenance, added downtime, and even greater losses. Fortunately, there are opportunities to optimize performance throughout the design and installation process—and lower the cost per kWh.
A thorough analysis of the proposed site and the associated electrical load is one of the first steps in system design. Poor decisions or miscalculations during these early stages can dramatically affect system performance. In many cases, system size is influenced mainly by budget and space constraints, with load requirements usually a lesser factor.
Squeezing in modules to increase array size can sometimes result in diminishing returns. For instance, at a partly shaded site, adding an additional module to a string may be more trouble than it’s worth. If even one module ends up with more shading than the others in the string, the whole string’s production can be reduced. Additionally, if an array is shaded above a certain percentage, some rebate programs reduce the amount reimbursed. Bypass diodes inside modules, specifying microinverters, and careful siting can help mitigate shading, but adding modules solely to increase system size beyond the shade-free maximum must be carefully considered to ensure that it is not done at too high of a cost per kWh.
A shade-free south-facing roof, in great condition, with a good pitch for the latitude and lots of space, is an installer’s dream. More typical, however, are roof penetrations (such as chimneys and vents) that obstruct and eat up potential space, trees that will eventually shade the site, and future buildings that may get placed on the empty lot next door. Some of these issues can be predicted by using one of the popular site analysis and shading tools available (see “Solar Site Assessment” in HP130 for details). Addressing others requires foresight and planning: Who will trim the trees? Will new construction nearby be tall enough to cause shading? Will the roof last the life of the array?
In most cases, the greatest amount of annual energy production requires the array to face true south and be tilted at an angle equal to the local latitude. This can often be accomplished with pole- or ground-mounted arrays. Roof-mounted systems are less flexible—typically the angle and orientation of the array will be that of the roof.
An array facing within 45° of true south and tilted to within 20° of the latitude will usually produce 85% or more of that of a perfectly oriented array. A program such as PVWatts (see Access) can be used to estimate array production based on the proposed installation. Calculating the output of the proposed array and dividing this result by the output for an array set at an ideal angle and orientation will determine what percentage of the ideal output the actual installation will generate (see “Shade-Free Solutions” in HP132).
Site variables may also have an impact on available rebates. Larger installations typically receive production-based incentives (PBIs), which are also becoming more common in residential programs. Under these incentive programs, payments are made over time to system owners based on the energy produced. Lost production equals a lower PBI payment. Other incentive programs pay rebates based on the system’s rated size. Rebates can be reduced if the array orientation deviates too far from true south, the tilt angle is too flat or steep, or if shading affects the array. The double whammy of lost production and lost incentives can dramatically increase the cost per kWh. (For more information on the impacts of angle and orientation, see “Optimizing a PV Array” in HP130.)
Off-grid and other battery-based PV systems may be the only source of electricity—or supply critical backup power—and proper design requires a detailed accounting of each electrical load to be powered. Load analysis tends to be more straightforward with grid-tied systems. For existing utility customers, one or more years of billing history is usually available, showing kWh consumed and total cost by month. Projecting the electricity consumption of new occupancy can be more challenging, and may require a thorough analysis of the appliances and other loads, along with an examination of utility bills from the previous residence, past tenants, and/or similar homes. The more data the better—if possible, view several years’ worth to take into account occasional aberrations, such as a holiday when lots of guests increased the load.
A thorough profile is necessary because of the various utility accounting methods for grid-connected PV systems, particularly with surplus generation. The most progressive programs use feed-in tariffs (a type of production-based incentive), which pay premium prices for RE-generated electricity. But most utility customers have only net metering—a kWh-for-kWh exchange. Net metering values PV energy at the retail price but usually does not pay for production beyond what the customer uses. Any excess production is usually carried over between billing cycles and for up to a year. After that it is lost—given away to the utility. In some states, the utility pays for excess generation at “avoided cost,” closer to the wholesale price per kWh.
Any time PV production is undervalued in this manner, the effective cost per kWh of the system increases. In contrast, many utilities have tiered rate plans, where the cost per kWh for utility electricity increases as more energy is used, especially during the hot summer months. During these times, increased PV output, whether by design or by the simple fact that there is more sun, increases the average value of each kWh generated by the system. In this scenario, a smaller system may have a lower price per kWh because a larger percentage of the energy generated will be credited at a higher rate—effectively “shaving” the most expensive kWh purchased during the year. Time-of-use rate plans further complicate the analysis (see “Time-of-Use Metering” sidebar). Coordinating between monthly consumption patterns, expected system output, and the ins and outs of specific utility requirements and programs can be daunting—just remember PV energy is pricey, so there’s no reason to give it away.
The list of available PV modules, inverters, and other equipment literally grows by the day. Ten-year inverter and 25-year module warranties are the industry norm, with some exceptions. Competing claims of performance and efficiency are complicated by standardized ratings that don’t duplicate realistic operating conditions (see “Realistic Expectations” sidebar). Cost, usually expressed in dollars per watt when pricing modules or inverters, always plays a role, along with availability and familiarity—look for the good deal, but in all cases, choose equipment that has a good reputation in the field.
Module Considerations. One of the biggest differences between modules is power tolerance, which is the expected variation from a module’s rated output. For example, a 200-watt module with a tolerance of +/-7% could actually produce from 186 to 214 W. Many modules are in the +/-3 to +/-5% range; some have a wider range, as high as +/-10%. Others guarantee a positive power tolerance, stating that the module will produce at least its rated power.
The wider the power tolerance in the module, the more likely it is that the PV array will have modules with different operating characteristics. For instance, the 200 W module mentioned previously has a power tolerance spanning nearly 30 W. Individual modules in this array are likely to operate at significantly different maximum power points. Because they are often wired together to an inverter capable of tracking only one maximum power point, the lower-output modules will tend to “pull down” the ones with higher outputs. When wired in series, the current of the string will be closer to the current of the lowest-rated module. When wired in parallel, the voltage of the strings will tend to be the average of the strings’ combined voltages. Since most grid-tied systems have source circuits wired in series and parallel, this results in lost power and means that a +/- power tolerance is usually only a minus. To combat this issue, modules with a narrow (or positive only) power tolerance can be used. Or consider using microinverters, where each module is paired with its own inverter or with inverters that can track more than one MPPT.
Matching the number and model of a PV module to an inverter requires careful consideration of local temperature extremes and the overall size and wiring configuration of the array (see “String Theory: PV Array Voltage Calculations” in HP125). Coupling an inverter to an array that barely meets the inverter’s DC input voltage will result in disappointing performance: In addition to the decrease in voltage due to temperature, modules will also lose some power output over time due to dust, degradation, corrosion, and increased connection resistance. Though it may take several years, these combined factors can result in a low-enough voltage to shut down the inverter. (See the “Grid-Tied Inverter Buyer’s Guide” in this issue for more information.)
Installation Considerations. Regardless of how well a system is designed, improper installation can result in poor performance. PV systems should operate for decades, and the materials and methods to install them should be selected accordingly. Wire, conduit, and associated hardware typically make up a small percentage of PV system cost—skimping on them may result in decreased output and, therefore, a higher cost per kWh.
Loose connections are a common and potentially serious installation issue. They lead to increased voltage drop, lost output, and added maintenance costs. At worst, the increased resistance leads to heat buildup and fire. Troubleshooting loose or sporadic connections can be time-consuming and frustrating, so minimize their likelihood from the start: All connections in the system should be tightened to the specifications of the device, and should be appropriate for the size and type of wire, as well as for the location. Torque wrenches are fairly common (torque screwdrivers much less so), but specs for tightening screw terminals are provided for components and should be followed.
Both PV modules and inverters operate more efficiently at cooler temperatures. While most grid-tied inverters are designed for outside installation and housed in outdoor-rated enclosures, they should not be mounted in direct sunlight, as this will cause them to operate less efficiently. In addition to the lost output, inverter life is likely to be shortened. While the expectations built into most PV financial modeling programs include inverter replacement, “burning” through several expensive inverters will dramatically increase the system’s cost per kWh. The LCD display in most inverters also can be rendered useless after too much sun exposure.
Placing overcurrent protection devices in excessively hot and sunny locations can also lead to unexpected downtime and the loss of energy production. Fuses and breakers are thermal devices—they rely on the heat generated by current running through them to “trip” and disconnect, protecting the wiring from overcurrent conditions. When operating in high ambient temperatures, the ratings of fuses and breakers are effectively lowered, meaning that they may “nuisance trip” even when carrying less current than they are rated for.
Until the fuse is replaced or the breaker reset, the output of the PV source circuit or array connected to that overcurrent device will be lost. Because grid-tied systems operate “silently,” the building will still have power even if the inverter is offline—lost production may not be noticed until the next electricity bill reports higher-than-normal usage. Keeping combiner boxes off hot roofs and out of direct sunlight and wiring roof-mounted arrays with home runs from each source circuit back to an inverter mounted in a shaded location (build an awning if necessary) can be good strategies to ensure that the system stays online.
Brian Mehalic is a NABCEP-certified PV installer, with experience designing, installing, and servicing PV, thermal, wind, and water-pumping systems. He is currently an instructor for Solar Energy International and works on curriculum development for SEI’s PV program from his home in Prescott, Arizona.
“Solar Survey,” by Justine Sanchez, HP130
“Optimizing a PV Array” by David Del Vecchio, HP130
“String Theory: PV Array Voltage Calculations,” by Ryan Mayfield, HP125
“Voltage Drop after NEC Requirements,” by John Wiles, HP80
Consumer Energy Center Equipment Ratings • www.gosolarcalifornia.org/equipment
PVWatts • www.nrel.gov/rredc/pvwatts