Internal conduit runs go through the attic rather than across the rooftop. These runs are easy with AC micro-inverters, but DC solar output circuits must be inside metal when inside buildings per code.
Many installers will stay outside the building, because attics are not fun to work in. But flexible metal clad cable can often be used to keep the internal cable run simple. It’s expensive and a specialty
item that requires advanced planning, but in many jurisdictions it can run directly from a rooftop transition box.
Figure 53. Rooftop Transition Box
The one pictured is a Soladeck and it can be useful just as an empty shell as a transition box. The MC cable then runs through the attic can comes out the soffit on the other side of the home, where the inverter is located out of direct sunlight. There are some additional conduit requirements on that other end, but ultimately the internal cable run through the attic results in a very nice looking installation. I use the same technique to move from one subarray on the roof to another in a visually appealing manner.
Figure 54. MC Cable and Cable Glands
The problem with buying generic material at the local big hardware store, is that it is not designed to fit nice and tidy underneath a solar array. Advanced installers pride themselves in clean cable management and well-looking array layouts. These additional steps result in long term resale value, whereas an unprofessional installation may result in a clunker.
Interconnection
Figure 55. Supply Side vs. Load Side Interconnection
Regarding interconnection, there are two main strategies. Either the power lines into the home are intercepted and tapped onto, between the meter and the main breaker, or the solar inverter is landed on a breaker on the existing electric service panel at the bottom of the panel. When planning a panel layout, it is also useful to reserve the top of the busbar for an emergency power supplies. In other words, instruct electricians to leave the top and the bottom of the busbar free of breakers when planning future expansion, with the bottom having higher priority.
At any rate, there are flexible options to interconnect an array
on site. The existing infrastructure can almost always be optimized or improved upon. But having a solar array by itself is much less complicated than adding a battery and wanting the system to operate and power the whole building, while connected to the grid, but switch into an isolated offgrid mode automatically. Those advanced project solutions are still being figured out today, and our other classes will better prepare you to make those decisions.
Backup Power
Figure 56. Battery Inverter Interconnection
But customers pursuing backup power should be prepared to add an additional $15k or more in project cost to incorporate a rather modest battery. The batteries in a 120kwh Tesla car, for example, costs Tesla about $15k for the batteries themselves. Taking into account the complicated interconnection equipment required, running a whole house off grid can add over $50k in additional project cost. So there are advantages of a grid connection, and much of solar design is planning the array around the customer’s specific rate structure.
I think adding a 200A automatic transfer switch ahead of a supply-side interconnection can make sense, although it is much cheaper and easier to stick to only powering smaller critical loads during a power outage for most customers. When planning an electrical room, it is not a bad idea to leave physical space around the existing panel for expansion.
Figure 57. Generator Interlock Switch
There are “low budget” manual ways to backup an entire home during a power outage using listed, inexpensive parts. Here is a manual generator interlock switch which can be used to provide generator power to a home during a blackout, and it could be almost as easily connected to the critical load output of a battery inverter (with additional considerations to prevent the inverter from feeding itself).
But that gets into the realm of solar design, and it turns out, we have a four hour class on solar design and computer-assisted design capabilities as a next step in your solar journey, so please reach out if you want more information.
Solar racking is fairly straight-forward, standard, and reliable.
There are positive attachments that lag screw directly into the rafter in most cases. The racking system is designed in manufacturer software to fit your roof and the built to order with some minor field customization.
No matter the roof material, there is a solar rack for it. There are multiple kinds of racking systems for just Spanish tile, ranging from older hooks to full tile replacement. But once the attachment point is determined, solar rail will run along the roof.
Figure 35. Positive Attachments
Module clips secure the module frame, the clip itself being bolted to the rail at just the right module height for a secure fit.
This piece is called an L foot, as it looks like an L. A bolt fits into this side channel and then a nut gets screwed down to clamp the
module frame between the rail and the clip.
As far as industry jargon, end clips go on the very end of the array and mid clips go between two modules. This little notch becomes a spacer in between two panels. There is some nuance in racking selection. I prefer racking systems whose rails have a wide U channel to lay cable into, whereas other installers prefer a more exposed wire which is better for rainwater run-off. That’s another discussion for another day.
Figure 36. Cable Management
But cable management is one of the more important jobs in solar installation and there are different tools for that job. Zip ties are useful for bundling cables together and keeping those cables away from pinch points and sharp metal edges of the rack itself. But there are also metal module cable clips that are worth having bags of because these actually accomplish a different goals of keeping the module MC4 connector whips tucked up underneath the module frame when installed.
Sometimes the clips are simply there to be an extra hand during a particularly challenging rooftop install. A zip tie is a more permanently secured solution. Both have their place on site.
Cable Management
Figure 37. Underneath the Solar Array
The overall goal of cable management is that you can look underneath
the solar array and not see any loose cables touching the roof or hanging down. A clean undercarriage is a sign of a good installer.
Array Skirts
Figure 38. Array Skirts and Snow Gaurds
Other material items to be aware of snow guards, which are
little clips dot the array, creating little grips for the snow to hold onto to prevent avalanches off of the roof. That’s only used in areas of very heavy snowfall.
These are array skirts, and they prevent pigeons, squirrels, and debris from getting under the array. Most installers do not install these but I recommend an array skirts, even a cheap one, if it is identified that leaves can fall on the roof, or other obvious signs that birds or squirrels are a concern. It is cheaper to install than fixing a problem afterwards.
Flashing
Figure 39. Flashing
The standard solar roof layout for shingles is flashed attachment points staggered across all the rafters of the roof, roughly 4’ on center. This distributes the load well and makes good use of the strength of the solar rail.
Figure 40. Lag Screws
Finding the rafters for the attachment points is part of the trade, but sealant and flashing works great on near misses. There are techniques to avoid unnecessary holes in the roof, but you do not want your installer to be someone who isn’t already comfortable fixing a hole in the roof!
There are some non-flashed systems that sit on top of the shingles, but they require more precision to install, and so are really for advanced installers.
Metal Roofs
Figure 41. Metal Roofs
These are clips for standing seam roofs, which are not the only metal roofing option. Positively attaching to a metal roof with screws sounds crazy, but it my favorite option, because a clip will only clamp to the top of that metal roofing panel. But a positive penetration that screws into the rafter is directly attached to the structure itself.
There are products for routing conduit across the roof, including metal roofs.
Rail-less
Figure 42. Rail-less Racking
This alternate racking design is called a rail-less and unlike traditional rail-based systems, it is commonly installed in landscape, which can look better or fit better under the right circumstances. I like this system when the location of the rafters cannot be determined, or the shingles are particularly weak, such as on cheaper roofing systems like foam-sealed trailer homes. But expert installers prefer it for cost advantages, despite the extra precision needed to install.
The rail on site can be useful for squaring the system, or as a stepping area on a slanted roof, or to position and handle the modules during the install. I recommend starting with rail-based systems.
However, there are even systems designed to attach directly to the roof decking, not needing the extra strength of a rafter attachment. So again, there are many solar racking options available for almost all roof types strong enough to support a few extra pounds of dead load.
Racking Engineering
Figure 43.Pull Strength
The racking manufacturer will provide engineering data, and sometimes that data can be very detailed, such as if they have their own sizing software available. For this particular mounting bracket, with the
safety factor of three considered, the allowable pull strength is 180 pounds and the pull strength of lag screw into the rafter is six hundred and thirty pounds. Yes the lag screw size and number of fasteners is important, but ultimately the decking has enough pull strength to hold solar in some circumstances, but not all. Attaching to the rafter is so strong you do not need to worry over it.
Figure 44. Live and Dead Loads
Staying in the interior of the rooftop, the racking software shows a force of a hundred pounds whereas if the edge of the roof in this other zone shows a 300 pound uplift force. Revisiting the pull strength of the attachment, the 180 pound uplift force (with a safety factor of three) for the decking is fine for the center of the roof, but not for the edges which would require a rafter attachment.
Advanced solar designers will consider this, and have attachment points closer together along the corners and edges of the array, but space things out to a maximum allowable span in the interior of the array.
Commercial Rooftops
Figure 45. Positive Attachment vs. Concrete Ballast
Many commercial racking systems cover the roof in concrete. Other racking systems positively attach to the roof. I’m an outspoken fan of commercial racking that gets rid of the concrete because covering the entire roof with loose concrete blocks does not seem like
a good long-term plan. A building’s maximum allowable load is often an assumed number, and the vertical load of earthquakes on this loose concrete has not been well studied.
These systems can weigh as much as 10 pounds a square foot
in certain spots and without the concrete they can weigh less than three pounds a square foot. So don’t be afraid of poking a hole
in a commercial rooftop – it’s something that can add structural benefit to the building and something a roofing contractor should know how to do.
Above the Ground Racking
Figure 46. Above Ground Mounts
Here are other kinds of non-metal racking systems, useful for metal-free rooftops or on landfills. Here is an alternate racking system for above the ground installation, for concrete pours. This can be useful for rocky terrain as well.
Screws + Piles
Figure 47. Ground Screws and Helical Piles
In rocky terrain or in other cases where foundation concrete is to be avoided, there is a ground screw which can be driven with a Bobcat with good reach and hydraulics. In sand or clay, a helical post can be a better fit. These mounts do not save any money compared to concrete but can improve logistics for the right site
Figure 48.Concrete-free Ground Mounts
Residential ground mounts are typically installed across two rows of support posts instead of one row, as you would find on a utility
scale project.
Figure 49. Fixed Utility Ground Mounts
The one post methodology is easier for maintenance but most residential ground mounts need the scaffold style foundation because the foundation depth of single post systems can be over the capabilities of a Bobcat.
racking system
Figure 50. Uneven Terrain
There’s different ways to level uneven terrain, ranging from
laser finders to string and a bandsaw.
Single Axis Tracking
Figure 51. Tracking
Single-axis tracking is a standard practice at the utility scale.
This is where the bifacial solar panels discussed before, the kind that collect sunlight underneath, have the most value.
Now the discussion will move away from solar modules and into inverters. One individual solar circuit is wired in series, with the modules plugging into each other positive to negative, head to tail. The industry labels that circuit a “string” of solar modules.
Solar arrays can be made of multiple strings of modules, and these circuits would wire into an inverter called a string inverter.
This is very traditional kind of system, with all the solar panels on the system being wired into one single inverter. Sometimes, ahead of the inverter, all the circuits are wired together and reduced down to a single cable. This process is similar to how power lines come into a house and distribute power on a breaker panel in that individual circuits are wired together in parallel and have individual fuses or breakers. This combining of circuits is parallel wiring, but most inverters today do any combining internally, such that the module strings are wired directly up to the inverter itself.
So a string inverter will manage multiple circuits of solar panels. Easy enough. At the commercial level, when the inverters are managing hundreds of circuits, they are called central inverters to denote them as larger than string inverters.
Shade and Safety
Figure 28. Shade Impact
Before we leave string inverters, let’s come back to that term maximum power point tracker. The solar panel produces the most power operating somewhere just under its maximum rated voltage and current. The inverter, being a computer, helps it determine those levels, but the system voltage and amperage is ultimately determined by sunlight and things like shade will directly impact it.
In the past, 10% shade could result in a 40% production loss. Having multiple power point trackers, such as module-level panel electronics, only 10% of the energy would be lost. This means more of the roof can be used as an array site.
Inverters might have one, two, or even a dozen “power point trackers”, which means that individual circuits can be individually controlled. This means the impact of shade is now confined to only one circuit. Although it is worth noting that a small amount of shade will still reduce the circuit power by 1/3rd to 100 percent, depending on the layout of the shadow.
But string inverters overall, with more powerpoint trackers, have become more shade tolerant, and can have arrays facing different angles, but only if all the panels on the same circuit or power point tracker face the same way and have the same electrical characteristics.
Back in the day, with only one power point tracker per inverter, all of the solar panels had to be uniform in circuit size and face the same direction.
Figure 29. Module Level Panel Electronics
Rooftops, as opposed to ground mounts, are subject to stricter fire code regulations, and string inverters by themselves are no longer good enough to be compliant with fire code. This is because individual solar circuits can reach 600V, or even more on a commercial building. But fire code wants the same circuits to be restricted to less than 90V on command. This essentially means that each solar panel on the roof must have a fire safety controller built into it.
This is a unique rule specific to the United States, and it drives up installation cost, and increases the need for accessible installation sites. Nonetheless, it is a safer design option and it is code.
Most rooftop installations will then take an additional step and have these module-level panel electronics also be voltage controllers. Optimizing the system voltage allows for longer circuits, and lets individually shaded modules to be bypassed from the rest of the system. This means solar panels that have particle shade from trees or chimneys no longer impact the rest of the array.
Micro-inverters
Figure 30. Early Growth in MLPE
Because they are code mandated, as well as useful, installing little boxes behind every or every other solar panel on a roof has become a standard practice in the United States.
There are two categories of these products. Starting small, we have micro-inverters. Every solar panel gets its own inverter and that means AC electricity comes out of the array instead of DC electricity, so general construction feels quite at home here.
Separately, micro-inverters are great for small projects. If you want to only install a couple solar panels, a micro-inverter circuit can be a great way to go.
Figure 31 Inverter Selection
The main problem with micro-inverters is that it is more expensive to buy a whole bunch of inverters instead of one big inverter and so the the any kind of module level panel is more expensive than a single string inverter, and the other category of Module Level Panel Electronics has a pricing advantage to micro-inverters, because it is closer to a string inverter.
So micro-inverters are great for beginners, and are used by many mainstream installers, but most solar installers who regularly install on rooftop projects prefer this other second solution.
DC Optimizers
The compromise solution between having every solar panel get its own inverter verses having one inverter for the entire system, is to take half of the stuff inside the inverter and put it up on the roof, for module-level voltage control, and then keep some of the stuff down in the inverter on the ground below. There is a slight cost advantage in doing so. Instead of DC to AC up on the rooftop, now the boxes just control the DC voltage and sent a steady voltage down to the inverter below.
DC optimizers may be the most popular rooftop architecture in the United States, but connecting all the wires correctly is the realm of a professional installer. While not terribly difficult, it is not as easy as wiring to a string inverter by itself. Micro-inverters, however are the easiest of the systems to install.
Just because a solar array is shade tolerant doesn’t mean that it will produce electricity if completely shaded. So ultimately, the total amount of sunlight available may depend on whether the site owner is able to cut down some trees. If shade cannot be avoided, such as a tall chimney, module level panel electronics are a great solution.
There might be a portion of roof that is sunny for eleven and a half months out of the year but for two weeks in December its shaded by the tree line and it might be economic to install an array there but even so, clients want to see their systems functioning all year round. So do not be reckless with shade analysis just because MLPE is specified.
With MLPE monitoring, every single solar panel can be checked in on, sometimes to the chagrin of the installer. But they fundamentally improve project safety and sometimes improve system production.
They allow for a safety switch, called a “rapid shutdown device” to be pressed on command and for the electricity from the circuits to drain to ground. This increases firefighter safety, such as if the solar array is burning the house down.
These rapid shutdown requirements have been on the market for some time now, and market leaders SolarEdge and Enphase have grown from startup to industry giants in a short period of time.
Inverter Specifications
Figure 32. Inverter Specifications
While on the topic, let’s look at a micro inverter specification sheet. This micro inverter is compatible with 60 cell modules
and that micro inverter is compatible with 72 cell modules at a higher voltage.
This output data is handy. It says “maximum units per 20 amp branch circuit” which is that these micro inverter units are wired into 240 volt double pole circuit breakers off the electric service panel. Sixteen micro inverters can be installed per branch circuit, so certainly when a solar array is 16 solar panels or less, a micro-inverter circuit is a good choice. The environmental rating is NEMA 6 which is industry-leading.
A string inverter specification sheet doesn’t tell how many solar panels fit on a circuit. To design with string inverters, a little more knowledge is required. The design work on a micro-inverter circuit is easier.
String Sizing
In string inverter design, the designer must ensure the circuit voltages do not exceed 600V. In other words, the designer must take the module short circuit voltage and make sure that number, multiplied by the number of panels on a circuit, does not exceed 600V at standard test condition, as well as at full power at the coldest design temperature. The actual calculation isn’t too complicated and there is a conversion table in the solar section of NEC for silicon modules.
But inverter manufacturers want to make it as easy as possible for you to design a system using their products and so they publish design software right on their websites for you to use. Many students wonder what the next step is after taking a solar training class, and my common answer is to start playing around with this free manufacturer solar design software. I do teach a 4 hour design course, where we have more time to spend on the capabilities of computer-assisted design in guiding decisions like string length.
Figure 33. Microinverter Specifications
But before moving on, let’s acknowledge that the same temperature considerations are given to micro-inverters, except because it is only one panel per inverter, this is a fairly easy item to figure out. Micro-inverter voltage is kept in range by identifying whether the micro-inverter is for a 60 cell or 72 cell solar panel, as referenced on its specification sheet. Then count the cells on the solar panel, and if it is what you are supposed to have, you are good to go.
Here is an example of a micro-inverter sizing tool, where the module
Specification sheets can be inputted and it’ll tell you whether or not that solar module is compatible with the micro-inverter.
At which point it is relatively easy to run a 20 amp branch circuit from the attic to the electric service panel. So if a pallet of solar panels has 26 solar panels, the micro-inverter circuit would need two 20A branch circuits, with up to 16 panels each.
When I shop for solar panels, I often shop for one or two pallets of modules, because that is how to get good module pricing on a residential sized project. Buying just a couple of solar panels isn’t going to get you a good price, unless you call a local installer, let them know your interest, and wait awhile.
Material distributors want you to buy pallets rather than an exact number of panels, and they will treat you as a newcomer with more respect if you make their job easier. I’ll go so far as to design my systems to be one pallet sized which allows material to ship directly to site or otherwise be easily received and handled.
Manufacturers some have 25, or 27m, or 29, or 30 solar panels in a pallet – it depends on the manufacturer.
Module Warranty
Figure 15. Module Warranties – Plural
Most installers will say solar has a 25 year warranty but in the same breathe state that they provide a industry standard one year workmanship warranty, so how do you accurately present the 25 year plus life of a solar array?
Almost all solar panels have a 25 year performance warranty but only a 10 year product warranty. Panels with full 25 year product warranties often cost twice as much as the same panel without one. But even the best panel warranty might only cover lost performance and not replacement labor. It can make sense to skip the premium today and use budget solar modules, rather than paying double for a technology that can become dated the minute it goes on the building.
The module performance warranty assumes that if nothing is going
wrong with the solar panel itself for 25 years, the cell itself will not substantially degrade or go kaput for 25 years. It’s a warranty of the longevity of the cell quality itself, like providing a warranty on the diamond but not the entire diamond ring.
What is not covered under the performance warranty is connection point of the cables coming out of the solar module itself, as well as how the frame seals the module glass, cells, and backsheet together, which to be honest is a pretty big deal, except that solar panels do last 25 years if installed properly. It’s that the manufacturer won’t vouch for the quality of the installation unless a much higher premium is paid.
Figure 16. Claim Procedure
How the panel is is handled from the time it leaves the factory to
the time it’s put on the roof, including the system design and the array location has an impact on how long a solar array can last.
Items such as abiding by clamp zones are important and skipping any critical instruction can result in potentially dangerous system failure. Instead, most manufacturers provide a 10 year warranty on solar module workmanship, so you have to watch out how you present the 25 year warranty. If the solar panel itself starts falling apart at year 15, but the cells are proven to still be functioning fine, that is not part of a standard 25 year solar warranty.
I recently came across a commercial project where the solar modules are facing straight up in the air, such that rain water will pool on top of the modules. It is an interesting design consideration, because pooling water is not good for the panels, although it is arguable that not much advantage is gained with the industry typical 5-10 degree tilt. A perfectly flat array on a flat roof has added structural advantages and could be the right design choice, but the manufacturer will not warranty the water tightness of the standard solar panel for more than 10 years. In normal circumstances this would not be an issue, but it could become a warranty an issue for creative solar projects.
Figure 17. Performance Degradation
As far as which is better between a bargain solar panel or the very best, I am pretty agnostic with an it’s all good approach. In very high-cost real estate markets or inaccessible surfaces it is worthwhile to go top shelf with an expensive solar panel and full 25 year module workmanship warranty. If real estate is cheap and site access is easy, a lower efficiency module that takes up more space for the same wattage is just fine.
Better performance warranties are broken out year by year, accounting for a little more degradation in year one than in following years. How that degradation naturally occurs has to do with how the electrons wear into the silicon as they start moving around. But if the installation is done with proper workmanship the expected life of an array is twenty five years or more. At that time, the solar panels should be producing eighty percent of what they do out of the box new, according to their performance warranty.
The performance warranty sometimes is 97% for year one and then a 0.7% percent decline per year from that. The actual performance loss is a little bit less, such that some warranties track very closely with actual module performance decline each year.
Another kind of “stair step” production warranty is for 90% of the rated power for 10 years and then 80% percent for the next 15 years.
This is a cheaper warranty than a “year over year” warranty, but it is often good enough for residential and commercial projects because when modules fail, more than 10% of it goes offline. The failure that I’ve observed is pretty obvious.
On commercial projects you can even buy 3rd party warranty insurance because that can be more reliable than trusting the actual manufacturer. My opinion is that if module-level panel electronics like SolarEdge or Enphase are mandated on rooftops, then defective modules are isolated from functioning modules, and so the module warranty matters less than almost any other value-engineered decision on the project. But more expensive modules with full 25 year workmanship warranties are available for the customers who have the budget.
Bypass Diodes
Figure 18. Bypass Diodes.
Let’s assume there is a solar module defect. Maybe it’s a soldered connection that came apart when electricity started to flow through it for the first time in the field. Maybe it arrived that way and the installer only determined during system commissioning. Maybe there are some leaves on the solar panel or some partial shading, but regardless when there is a failure inside the circuitry of a solar panel, there are bypass diodes so that part of the solar module can be damaged but electricity still flows through the other parts.
In that circumstance, the solar production would reduce by 1/3rd if there are three bypass diodes. Module voltage or amperage might read exactly 2/3rds of its neighbors during a failure event. It is very unusual to deal with module performance warranties. They’re usually defects that apply to the 10 or 25 year module workmanship warranty, rather than the 25 year performance warranty.
Shipping and Handling
Figure 19. Module Handling
Most module issues have to do with shipping.
More on those shipping defects. Some solar panels come in boxes on pallets and others come outside a box in wrapped stacked bunches. In particular, the latter can bust pallets if the shipper takes a fast turn. Likewise a forklift can spear a pallet and cause damage.
There are a few things that installers can do to damage a solar panel on a pallet, so its worth pointing them out now.
Modules can bust by leaning them against a wall for the wind to blow over. More insidiously, modules can break on site or during shipping if tools or other items are stacked on top of stacked modules.
What installers can do to prevent their end of workmanship defects is not to place heavy objects on the module surface itself. Don’t put elbows or knees, or tool boxes on the panels. If needing to Spider Man
across a solar array to access a particular nut or bolt, the module frame itself can be put to use and anything to distribute your load across the frames rather than the glass, the better.
Scratches and Micro-Cracks
Figure 20. Module Scratching
It is also easy to leave a noticeable white scratch across the module glass when lifting solar panels up off a pallet, particularly when this is a one person job. If modules must be slid rather than lifted, such as to get a better grip before lifting, care should be taken to slide the modules along their frames, rather than having a frame drag across the glass of a module below. This happens when taking a module off of a pallet.
So don’t scratch the glass and keep your weight on the frame of the panel whenever possible.
This section of the panel can take one inch hail at 40 mph, as well as high evenly distributed loads, but a point load large enough in the wrong spot, such as sitting or leaning against a ground mount after it has been installed, can damage the otherwise robust panels – even in ways which can’t be seen or detected for years to come.
Snail Trails
Figure 21. Snail Trails
Little micro-cracks can occur within the solar panel which are not obvious and only become problems later. Electricity flowing over these micro cracks like water carving into canyon, can create larger defects by unzipping the silicon cell in these lines called snail trails.
Sometimes these snail trails are problems, and other times not. Sometimes they are caused by manufacturing defects, other times environmental or installation error.
All-Black Panels
Figure 22. All Black Panels vs. Standard
My favorite solar panel is an all-black panel, with a black frame and black plastic back sheet instead of a white backsheet that hides the individual cells visually, especially from a distance. That look is where most of the value of a residential solar array actually comes from, which is why I almost always use black solar panels on residential jobs. There is a range of all-black panels, the lower end ones having exposed grid lines, the more expensive ones being uniform sheets of black glass, like an infinity pool for your roof.
Frameless
Figure 23. Frameless
For the sake of variety, these are frameless solar panels. Potentially, frameless solar panels could take over the industry. The module frame itself, being metal, is a weakness. But I would put frameless solar panels well within the experimental category at this point, although they are widespread enough to have racking options available. Most frameless modules today have glass fronts and glass backs, which make them heavy and easy to break. Perhaps a farming environment with heavy ammonia gas or an area with heavy lightning would benefit from a metal-free solar racking system panel.
Bifacial
Figure 24. Bifacial
This solar panel has a glass front and glass back, and is making headlines in the utility-scale sector. This is a glass-on-glass or bi-facial solar panel, and when installed off the ground such as on a utility-scale tracker, they can increase project cost-effectiveness by a whopping 12%, because they can increase production by 35% using similar real estate. Expect to see this market expand in the future, as companies like Dupont have translucent plastic module backsheets which can replace the glass. Again, these products are widely available but small enough to be still considered as novelty. The benefits of adding such modules onto a roof are reduced – I think the glass on glass ones are too heavy for slanted rooftop installation but can have a great building-integrated PV effect.
One easy way to remember the added benefit of these kinds of panels is that the underside of a solar array is by definition shaded. A solar array on an overcast day will generate 10-30% of what the array would do on a sunny day, and a bi-facial solar module will increase performance by roughly that amount, depending on how much space there is between the module and the mounting surface.
Solar Shingles
Figure 25 Solar Shingles
Solar shingles are available, although one of the largest manufacturers of solar shingles (no, it isn’t Tesla but roofing manufacturer CertainTeed) makes solar panels as well as solar shingles, so what does that tell you about the confidence in the product class?
Solar shingles look nice from above, but ugly from the ground. All the problems solar panels have looking good on uneven surfaces are made 10x worse with solar shingles. And even when installed correctly, I do not think they look any better than a well designed traditional array.
While it is uncertain that a solar shingle array will look better than a traditional array, it is guaranteed to cost substantially more, both in installation and maintenance.
Separating Material + Labor
Figure 26. Distributors
Solar panels, at a minimum, are best bought by the pallet. If seeking to buy just one solar panel, call your local solar installer and sometimes it is convenient for them to sell to you at cost. But when evaluating a residential project, one pallet of solar is a great starting point.
To get pricing, it is easy enough to find solar distributors online, and if you purchase by the pallet, it is likely they will sell to you, despite being a new customer. Look for newsletters to sign up for, or contractor programs. This is not as rigid a network as in other more established industries such as air conditioning. Anyone can buy solar panels for around the same price as the mainstream installer if they stick to pallet purchasing and do a little bit of material bidding.
Add a bit of design and project management work, combined with some hourly labor, it is a relatively straightforward process to manage a solar project and have it come in 50% less than turnkey installation price. In other words it is useful to get multiple bids and evaluate each for its merits, rather than just focusing on cost.
Welcome to our solar training program. Solar power is very scalable and many lessons in residential material selection also apply to commercial projects and utility scale, so this is as good a starting point as any to increase your knowledge of the solar industry.
Feel free to ask questions and use the chat widget to communicate. In the next two hours we’re going to review modules, inverters, and solar racking terminology as well as balance of system material.
We’ll also talk about supply chain, such as how installers purchase material for small projects. This class will stay away from design scope. There are other classes. We offer that cover that which you can ask about, but let’s get started with this.
Solar transitions our grid to site-generated clean power which has many values. The more knowledge you have up front the better of the result will be.
Modules, Panels, Cells
Figure 2 Modules, Panels, Cells
The solar industry calls these solar modules, instead of solar panels, and we will use the term interchangeably in class.
But technically we call them solar modules, as electric code already uses the term panel to describe other important things, and also the term module does mean something specifically.
An electric service panel holds the breakers that distribute power to a home, including the solar array. A roofing panel can be a section of roof that a solar array sits on top of. So the term solar panel is ambiguous.
Figure 3 60 vs 72 cells
The term module in code is used specifically to refer to a grouping of individual cells. You might have heard of a battery cell. A 12V battery might be a single circuit of six 2V cells. Similarly, a solar module or panel contains a grouping of individual cells.
Most residential panels are 60 cells and most utility scale panels are 72 cells. Counting the cells here there are one two three four five six and then along the length there are twelve. Six times twelve is 72 so this is a 72 cell solar panel.
This solar module has a black backsheet instead of a white backsheet but a trained eye can still count the cells. We still have six cells across but this module is a little bit shorter than the 72 cell module with only 10 cells down.
That makes a 60 cell module and the the difference between a residential 60 cell panel compared to the utility-scale 72 cell panel is these extra two rows of cells, which give the module a little bit more height, weight, and voltage.
A larger panel means slightly more bang for your buck all the way throughout the whole project, but the problem with 72 cell panels is they’re a little too tall and heavy for one person to safely handle on a slanted rooftop.
So residential rooftop solar will have the 40 lb, 3.3’ by 5.3’ 60 cell solar panels typically, again because they are easier to manage on the roof.
On ground mounts, or flat commercial projects, the larger 72 cell panels have slight cost advantages. A few specialty module manufacturers might deviate from these standards as well. 72 cell modules are allowed on residential roofs, they’re just a little harder to manage.
So a solar panel is about 3 and 1/3 feet by 5 and 1/3 feet. Maybe a little different but that’s an easy size to remember. It’s just small enough for one person to pick up and walk around with.
The 72 cell panels used in utility-scale projects are 20% larger, with 20% more cells. They weigh 20% more and have 20% more voltage too. So a 72 cell panel can be over 6 feet tall and that gets a little hard to manage up on a roof. A 60 cell solar panel weighs 40 pounds and so a 72 cell panel weights 50 pounds.
Mono vs. Poly Silicon
Figure 4 Poly vs Mono Silicon
Silicon for solar panels is grown as a rectangular sheet, known as polysilicon. Or it is grown in a cylinder and sliced like a salami called monosilicon.
Whether they are cutting the rectangle of silicon or the cylinder of silicon, the interior of the silicon is more pure than the edge condition of the silicon crystal, and now manufacturers sort the more efficient cells onto higher efficiency panels efficient panels. Cell sorting is one of many reasons solar panel efficiency has increased dramatically over the years.
Less efficient cells go on to the less efficient panels which is why a module specification sheet will have the specifications for multiple solar wattages, all which have the same dimension. The cell sorting creates all these different wattages out of the same production batch and they are priced differently as well.
Mono crystalline solar is slightly better than polycrystalline silicon, due to these slight differences in the manufacturing process. But they’re like siblings born in the same year, in that they are very similar yet not quite twins.
Futuristic thin film solar panels, such as translucent solar panels do exist, as do solar shingles, but all that stuff is a very small section of the market. Silicon is the clear market leader right now at a couple bucks a pound.
On a side note: if a solar panel weighs 40 pounds, and silicon costs $1 per pound, then there is approximately $40 worth of silicon in a $150 solar panel.
Module Efficiency
Figure 5. Module Efficiency
Solar module efficiency is the percentage of energy that hits the surface of the panel versus how much electricity flows out the backside of it. An impressive 20% of natural sunlight can be converted to electricity at its standard test condition. That number continues to climb.
Today’s solar panels capture almost twice as much electricity as the technology fifteen years ago, solar panels which are still producing today. Scientists themselves have underestimated how much electricity can be squeezed out of annual improvements in solar module design.
For example, a recent advancement in silicon is called PERC, which makes the bottom of a solar cell more shiny. This additional reflection increases total solar panel efficiency by 1%.
But let’s get back to basics with the specification sheets for now.
Module Specification Sheet
Figure 6. Module Specifications
Let’s take a closer look at module specifications. The module power rating is how many watts of power the panel will output under certain lab test conditions. This information can be used for pencil and paper design but nowadays most design decisions are guided by computer design software.
This solar panel is an older model so it has a slightly lower wattage than most solar panels being installed today. Most 60 cell solar modules today are around 350 watts, just under 10 amps and 40 volts.
MP or MPP stands for maximum power point. Basically the maximum power point is the operating voltage and amperage, rather than the maximum voltage and amperage. The maximum amperage is the short circuit current, and the open circuit voltage is how high the voltage would be on the circuit if it were fully energized without being under any load at all, like a light switch turned off with voltage on one side of the switch, and metal connected to ground on the other, with nothing but voltage potential in between.
The maximum voltage and amperage are used in system design, rather than the operating voltage or amperage. A solar array may be designed for 600V, but will operate at a lower voltage.
All the figures given on a module specification sheet are listed for a standard lab test condition. Just like a car gets a mile per gallon rating, the solar panel gets a wattage rating at a standard test condition.
But solar performance changes hourly, as well as seasonally, as the sun changes position in the sky. Overcast days diffuse the sunlight, and only sometimes is this enough to turn the panels off.
Or the numbers can jump around on a partly cloudy day. So the numbers on the spec sheet are just a starting point, and how might they compare with solar performance on a normal day?
The standard test condition amount of sunlight is 1000 watts per meter square. There’s a free public software called PVWatts and as an example I’m going to pull up Biloxi MS.
Figure 7. Analyzing Environmental Conditions
This is a PVWatts example for a 1kW in Biloxi, MS. Like other coast lines, Biloxi gets better production than further inland because clouds tend to blow over rather than stay still.
Anyway, at the end of PVWatts, the hourly results provide an estimate of the actual amount of sunlight that will hit the surface of the solar panel. PVWatts takes into account array tilt and orientation, and for this 18 degree tilt, south facing array, the array irradiance is listed on the spec sheet as the lab condition of one thousand watts per meter square in Biloxi is equivalent to high on a cold spring day.
Compared to this winter day in PVWatts, the solar array only gets 360 watts per square meter of sunlight. That’s only 30 to 40 percent of the sunlight used in the lab test and so the solar array only produces thirty to forty percent of its rated capacity on this cloudy winter day. On a sunny day, still in winter, the array gets up to 640 watts per meter squared. Moving into July, temperature begins to impact the array by 10-20%, even if it had full exposure to a thousand watts per meter squared sunlight.
So calculating solar performance isn’t as simple as taking the solar array size, multiplying by hours of daylight, and getting the amount of energy the solar array will produce from it. The wattage rating on the module assumes 1000 watts per meter square, but it rarely gets that amount of sunlight out in the field. A software like PVWatts is needed to estimate solar production. It’s easy to use and covered in other classes.
Figure 8. Rooftop Temperature
Then there’s this other part of standard test condition called the module temperature and that’s done at a 25 degree Celsius module temperature. How does that compare to rooftop temperature? 25 degrees Ceclius is equivalent to 77 degrees Fahrenheit, so how hot does it get on the roof when it is 77 degrees out?
Again PVWatts comes in handy. It provides ambient temperature column and module temperature, and it shows how to convert ambient temperature into module temperature.
But standard test condition is done at module temperature, and so standard test condition is performed on the equivalent of a 60 degree day outside, which is not very typical weather for most of the year.
Temperature Impact
Figure 9. Temperature Coefficients
Temperature Coefficients
On the module spec sheet, there are temperature coefficients to show how voltage, amperage, and power are impacted by temperature.
For example, if the rooftop is 50 degree Celcisu and standard test condition is 25 degrees Celsius, multiply the difference (25 degrees) by 3.4 percent and that’s a 8-9% loss of voltage due to increased heat.
The temperature coefficient of amperage has a positive sign. But the temperature coefficient of voltage has a negative sign. Whereas voltage decreases with temperature, amperage increases with temperature. That’s worth pausing to memorize, but if you are having
Problems understanding voltage and amperage, I recommend starting with a volt times an amp is a watt. In other words, power is the multiplication of voltage and amperage. Voltage and amperage are different things that make up power, and we will leave it at that for now.
Bringing us back to our temperature coefficient discussion, the temperature coefficient of power is a -0.4 percent so at 50C, there is a 25C difference with standard test condition.
A 25 degrees temperature differential times .4 correction factor percent means 10% of the power is lost due of temperature alone.
Sunlight, also known as irradiance or insolation, has a linearly proportionate effect on solar module power. If the air has twice as much sunlight in it, a solar panel will produce twice as much energy minus any temperature effects or starting energy. This is our open circuit voltage with all voltage and no amperage.
The short-circuit current is the opposite, with all current and minimal voltage. The bend in this curve here is the maximum operating power of the Solar panel, with the right level of voltage and amperage to produce the most power under normal operating conditions.
So the panel will operate at a little bit less voltage than its open circuit voltage, and a little less Amperage than its short circuit current.
Load vs. Voltage
Adding any electrical load to voltage will reduce system voltage,
Because electricity is not infinite. Too large a load on too small a battery will lower the battery voltage to the point where it can’t deliver any power. Too much load on a grid can lower the grid voltage too.
But normally, our very large grid provides us with stable voltage because our loads are small and the grid is large. The solar panel too has an ideal operating voltage and amperage, and in the past, this would lead to a difficult discussion about keeping solar panels away from shade. Maximum power point tracking will reappear in the inverter discussion.
An Outdoor Product
Figure 10. Solar Module Features
Solar panels are designed to go outside. Manufacturers hail test them at 1” thick at 45 mph, similar or tougher than most roof material.Installed correctly, solar can strengthen and protect the roof. But solar isn’t bulletproof.
It is strong and durable enough to give a false sense of security. It’s a quality object, something you can throw in a truck and drive down the road. By itself it will not likely break or shatter. If you drop it flat on its face, it might even be okay.
It’s durable and designed to face the elements. The specification sheet also reveals physical load data such as snow load and wind rating. The panels are tested for hail impact. They might have a positive or a plus and minus power tolerance which has significantly improved over the years.
All of these numbers are fairly similar throughout the industry. A particular manufacturer might have a patent on one unique ability, but the features are more often standard than special.
Some manufacturers provide different module frame thickness to choose in hurricane zones. Sometimes this information can only be found in the manufacturer installation manual.
Module Clamp Zones
Figure 11. Clamp Zones
Here’s an example that illustrates various ways to install this particular manufacturer’s solar panel to a rack. It provides different wind load and snow loads based on how the where the module clips are installed relative to the module frame.
Also shown is a technique of adding an additional rail for more reinforcement, such as to meet hurricane wind load requirements.
MC4 Connectors
Figure 12. MC4 Connectors
There’s two dozen mainstream solar manufacturers still in this industry so by and large module quality is very good. The type of frame, the color of the frame, the gauge of wires coming out the
back of the panel, all of that information can be useful to know. One of the notable industry standards is the MC4 connector coming out the back of the solar panel.
A common point of system failure is the field-made MC4 connector, which creates the home run circuit connection between the array and the inverter. Even when doing things correctly, it can be easy to confuse the male and female housings and metal connection tips. Furthermore, the connector itself should not be exposed to any direct rain (such as dripping down in between modules) or accumulating water. So attention must be paid with cable management in locating the connectors, in addition to making them. Finally, they click together with a nice snap.
Figure 13. Module Dimensions
Even the information like how much length comes off the back of the module for that MC4 connector plays a role in advanced solar design, such as how to route the cables underneath the array. This is the
stuff that like solar installers take pride in, having the extra slack tucked out of the way in a workmanlike manner, as to not attract nuisances like squirrels or debris.
Since we are starting with material selection, a more advanced material selection note is to review the voltage rating on these MC4 connectors. Mc4 is the connector standard in the industry, and some field made connections meet up with the plugs that come installed on the back of the solar module. While these connectors are universal, some are not rated for 1000V commercial systems.
The detail in a module spec sheet can assist racking planning. some manufacturers will hollow out this piece of metal in the module frame
To save a little money. The US manufacturer SolarWorld does this with some of their panels. On standard solar, the robust module frame allows the racking underneath to run in portrait or landscape orientation and so the solar panels can be designed in portrait or landscape orientation, but not with these hollowed out module frames.
