![[Image of Man cleaning PV array]](f_shade.jpg)
![[Image of tracker]](f_track.jpg)
Tracking is using some sort of mechanical system to keep your solar panels
aimed directly at the sun. There are two main types of tracker. The
first, called "single-axis trackers" turn panels in an arc that follows
the sun from horizon to horizon each day. The sun also moves about
40 degrees up and down in the sky with the seasons however. The second
type, called "two-axis trackers" both follow the sun from horizon to
horizon AND follow it from solstice to solstice.
The single axis tracker gets you about 12% more energy output than stationary panels at the same tilt.
Dual axis trackers get you about 25% more energy output than stationary panels.
Trackers have some drawbacks, of course. Most of them use some power to track the sun. That power will reduce the benefit of tracking by the amount of power it takes to track. They are moving mechanical devices and thus are subject to wear and maintenance. That energy and expense must be planned for. The biggest drawback is that accurate and reliable trackers are quite expensive. It takes a long time for the energy gain to pay off the expense.
I do not have hard numbers. According to the books I've read, trackers do not generally make economic sense.
I invite persons with more information on trackers, pro and con, to contact me.
Even worse, the energy created by the illuminated cells has to go somewhere. Guess where is left if it goes out of the illuminated cells but doesn't go out of the series of panels? That's right. It gets dumped as heat out of the shaded cell. Often the heat is high enough to damage the cell. It is for this reason that PV arrays need bypass diodes. See the system components discussion below.
Even with bypass diodes to prevent damage to system components, shade means dramatic loss of power since even a single shaded cell usually means an entire panel is bypassed. For this reason it is very important to keep your array unshaded.
Many sections of the NEC come into play in making a solar powered home, but as of 1993, the code contains an article (article 690) specifically about solar photovoltaic systems.
The major difference is that SA systems require copious energy storage for periods when the sun is not shining (night, for example!) and thus all the support equipment that goes along with those batteries. SA systems are far more expensive and require considerably more maintenance than UI systems. Ironically, the SA systems are often more economical because they tend to be used in remote applications where grid power is prohibitively expensive.
Let's outline to components of an SA system and then we can talk about which elements are not needed for a UI system.
The PV Panels are the most obvious system component. We've been talking about these for most of this document. There are a few items to talk about here that will help make the planning section easier. First, we already know that voltage is the average energy of the electrons in a circuit. We know that current is the number of electrons passing a point in the circuit every second. As you may know from your home wiring, you have to use wire of a certain size based on how much current the wire will carry, not how much voltage. (A 50 amp circuit for an electric furnace requires much heavier wire than a 15 amp circuit for lighting). This is because the motion of electrons through a wire heats the wire. As it heats, it's resistance grows so the flow of current heats it more and so on. If the wire is too small, this effect will cascade until the wire melts or causes a fire. Remember that power is the product of voltage and current. Because of this, if you can get the voltage of your solar panels up, you will reduce the current and still have the same power. Fortunately, there is a way to do this. When you wire cells in series (meaning the plus of one cell goes to the minus of the other and so on), the voltages add together. (As an aside, when you wire them in parallel, the current adds) Panel makers wire cells in series and in parallel to get fairly reasonable voltages and currents. Even so, you might want or need to wire panels in series in order to get the voltage up (in larger systems) to 24 or 48 volts in order that you don't have to string heavy cables all over the place. Also, since there is variation in the output voltage of panels, you should carefully measure and match your panels. A series of cells will only put out as much voltage as the weakest cell in the series. That weak cell will dissipate the excess from the others inside itself as heat. This can be a very bad thing. For that reason, you cannot simply mix panels together haphazardly. Keep both of these effects in mind. They are important in many aspects of PV system design.
![[Image of Solar Shingle]](f_roof.jpg)
PV panels are even available in the form of roof shingles! Photo from
NREL
The support structure (and optional tracker) is another fairly obvious system component. The panels must be oriented towards the south and mounted at an optimal angle. Now we get into a difficult question. If you have a fixed array (non-tracking), what angle do you put it at? If you put it at an angle equal to your latitude you will get optimal output on the equinoxes. If you put it at your latitude plus 23.5 degrees (the lowest angle of the sun -- midwinter solstice) you will get your maximum output in the dead of winter. If you put it at your latitude minus 23.5, you will get your maximum output during the summer solstice. Of course, the days are short in the winter and long in the summer, so that affects the figures as well. For the most part, the greatest concern in an SA PV system is to obtain uniform system performance. It turns out that latitude plus 15 degrees serves to get the best mean performance with minimal seasonal variation. (See Strong, pp. 67-71) This is the optimum angle for a stand-alone fixed array.
For a grid intertied system where maximum power production is your goal, you should tilt your array at an angle equal to your latitude. You will get a fairly large seasonal variation in output, but you will get the maximum power output for a fixed array.
The optional tracker helps to eliminate the above problem. Trackers come in two flavors. "One-axis trackers" follow the sun from horizon to horizon during the day. This can increase the power collected by about 12%. "Two-axis trackers" both follow the sun horizon to horizon, but also follow its declination as the sun changes angle through the seasons. This can raise the power collected (as compared to fixed mounts) by about 25%. The trouble is that these systems are expensive and most consume power and, as always, moving parts add maintenance costs and time. I have not seen a full analysis that would determine once and for all how long it takes for tracking systems to be cost-justified, or if they ever are. I can say that a lot of folks who do PV don't add this expense to the already substantial start-up costs of a home PV system.
![[Roger Edwards lightning photo]](f_light.jpg)
Lightning protection is a must. There are a host of devices for
this purpose, from glow discharge tubes to the more common metal oxide
varistors. Basically, these devices will short the array to ground
(and it has to be quite a good ground!) in the event of a high voltage
transient (usually a nearby lighting strike). The purpose of these
devices is to protect the devices "downstream" from voltages that
might cause fires or explosions. While direct lightning strikes are
mercifully unlikely, even nearby strikes can cause tens of thousands
of volts to be conducted or induced in system components. These
devices generally "kick in" at 50-350 volts (depending on the size
of the system) and they short to ground, thus sparing the expensive
downstream components the brunt of such a massive electrical discharge.
These devices will let a system survive the majority of lightning
induced transients. Nothing is going to prevent some damage, however,
in the event of a direct lightning strike. Luckily, as we said, such
events are statistically rare.
The "string combiner" is a device invented for PV systems. This is a box in which the terminals of a set of panels connected in series are joined together through "bypass diodes." What is a bypass diode? Well, remember what we said about how a series of cells is only as strong as its weakest link, and how we said that this weak link dissipates the excess from its neighbors as heat? If a panel goes bad or if it gets shade, its voltage drops and it becomes a load. PV cells aren't particularly good at dissipating heat, nor do they take excess heat well. Remember that a diode only allows current to flow in one direction? The bypass diode prevents current from flowing backwards into a panel. Effectively, it cuts the panel out of the circuit. This is not desirable in that the output of the array drops dramatically, but it is vastly preferable to the expensive damage that might otherwise occur. The string combiner is not always a separate and descreet component at this time, but as more PV installations appear and electrical inspectors and insurance companies get more savvy, expect them to require tested, listed "string combiners" in PV systems.
The voltage regulator in a PV system does the same thing as the voltage regulator in your automobile. It controls the voltage that goes to charge the batteries. The regulators used in PV systems are quite a bit more sophisticated than those in your car however. The most glaring difference is the presence of either a relay or a diode that "blocks" between the batteries and the PV array. Why? When the sun is shining, the voltage of the PV array is higher than the battery voltage and charge flows into the batteries, just as it should. When the sun goes down, the voltage from the PV array drops to a very low level. At this point, the batteries would start to discharge through the array, making the array hot and dumping all the charge in the batteries. This is not the use to which we want to put the day's energy! There's a lot more to know about regulators, but we'll defer that for now. There are very sophisticated voltage regulators on the market that are called charge controllers. These often have battery temperatur sensors, variable voltage output, and programmability for different types of batteries.
The batteries have the rather obvious job of storing the solar energy accumulated during the day for use at night or when the sun is blocked. Batteries are a big topic and they have their own section here in the FAQ.
A stand-alone system is likely to occasionally encounter conditions where the battery storage is insufficient to cover an extended sunless period. Such systems will require some sort of backup generator. These will be used to keep the batteries at charge. Backup generators are available powered by gasoline, diesel fuel, natural gas (methane), and liquid propane (LP) gas.
Since most backup generators are designed to produce AC power, a battery charger is likely to be needed to provide DC for battery charging.
The inverter (which could be considered optional if you convert to DC appliances) is used to convert the DC output of the batteries to AC for use by machines, tools, and appliances.
A utility interactive system can do away with the voltage regulators, batteries, chargers, and backup generators. Some folks keep all of the above for the psychological value of being able to ride out grid failures. Your actual independence is always illusory, however. No matter how long PV may last, your generator and batteries will not last forever. PV is not and cannot be a way for us to cut ourselves off from the rest of civilization.
This might make excellent economic sense for homes located in inaccessible places or a long distance from existing grid service. Stand alone systems require batteries to supply electricity when the sun isn't shining. Often it is also necessary to have some sort of fuel-powered backup generator to recharge the batteries when multiple overcast days occur.
The utility interactive PV installation typically does not require batteries. The utility grid is used when the sun isn't shining and excess power production is sold back to the utility when it is shining. Some systems require a single battery to keep the inverter's "brains" working if the grid experiences a power failure.
A utility interactive system might have batteries if the ability to ride out power failures was desired.
Despite the many variations on a theme charge controllers come in two basic flavors; Series controllers and Shunt controllers.
Both types protect the battery from being overcharged. While PV systems can work with either type, most wind and hydro systems require a shunt controller.
Series controllers are more popular since as a rule they are cheaper for a given current capacity.
Series controllers work by breaking the connection between the charging source and the batteries when the voltage reaches a predetermined point. This occurs when the batteries are fully charged. Any additional current supplied to the battery after that point only results in water loss thru electrolysis.
Shunt controllers work by applying additional loads on the battery bank to keep the voltage from rising too high when the batteries are fully charged.
Controllers are rated by the maximum current they can handle and the nominal voltage of the battery bank ie; 12 or 24volts. Normally the voltage from the charging source is comparable to the battery voltage but some specialized models allow a much higher voltage on the source side than on the battery side. Normally this is done to reduce power loss when the battery bank is located a substantial distance from the charging source.
Another important fact to remember is that most controllers are designed for a specific battery type ie: Gel-Cell, Flooded-Lead-Acid or Nickel-Cadmium, and cannot be adjusted to work with other types. Using the wrong type controller can result in excessive water usage or under charged batteries. Both conditions can result in reduced life expectancy for your batteries.
PV panels put out DC, where the voltage is constant and the polarity does not change. Since most household appliances are not designed for DC a device is needed to convert from DC to AC. This device is called an inverter.
The problems (which I promise I'll mention below!) are slightly reduced by a second type of inverter, the so-called "modified sine-wave inverter." The modified sine-wave inverters add a "rest " at zero for a short time on each transition. This makes the wave form look slightly more like a sine wave. Examples:
Square wave: Modified sine-wave:
______ _
| | | |
| | | |
| | _| |__ _
| | | |
| | | |
|____| |_|
(Each of these is one cycle, or 1/60th of a second)
I have been told (but I have not researched or verified this) that there is another technique that is sometimes called "modified sine-wave," but is more strictly called "pulse width modulated sine wave" inverters. These work by sending a series of positive-going square waves, then a series of negative-going square waves, but varying the width (or frequency) of each pulse from short to long to short positive-going, then short to long to short negative-going. The mean of these is a sine-wave.
As I said, I have this information only anecdotally. This is still a rather harsh waveform, but it is better than either of the two I've drawn above. I am looking for good information on all of these, especially a list of which inverter manufacturers are known to use which technique. I would then contact those manufacturers for more information.
Many solar advocates strongly endorse converting lighting and appliances to DC. This makes sense in that an inverter wastes some of the incoming energy (as much as 30%) and that requires more (expensive) solar panels to offset the loss. Realistically, however, most of us will want to go on using and buying whatever appliances we want and since most are AC, most systems will require at least one inverter.
The following question was submitted to me by an e-mail correspondent:
BTW, some say that modified sine wave power inverter could slowly fry sensitive equipment, is that true?
A sine wave is, mathematically, a pure signal. It contains only the fundamental frequency. A square wave is, mathematically, the sum of the fundamental and ALL of it's odd harmonics. In other words, it's a bunch of sine waves of different frequencies added together.
AC motors, power supplies for electronics (computers, TV's, stereos, etc.), ballasts for fluorescent lighting and so on all contain "inductors" (coils, windings, transformers, etc.). Inductors have a "preferred frequency." (This is a gross simplification. I invite someone with a stronger EE background to fix this up!) Frequencies higher than this preferred frequency will not pass through the inductor but the energy they carry has to go somewhere. It turns into heat (and often, noise!). All of the devices just mentioned will hum loudly when fed with square wave or modified sine wave power. They will also become hot quite quickly. Many of these devices will not be designed to withstand this excess heat and some may fry quickly rather than slowly. AC motors will both get hot and be slower to start and much weaker (because much of the incoming energy is turning into heat instead of motion.)
Devices that are not loaded with inductors (incandescent lights, toasters, electric stoves and ovens, etc.) will not care a bit about the incoming square waves. These are "resistance" loads. There are also "capacitive" loads, but I'm hard pressed to think of any household appliances that are primarily capacitive. (Actually, all loads are to some extent inductive, resistive, and capacitive -- we simply use whichever effect they use most to name their type.)
When you have a grid intertie system, one of the limits that will be set by the power company to which you connect will be the total harmonic distortion they allow. This "misshapen" wave is what they mean when they refer to harmonic distortion. Usually, the amount they allow is in inverse proportion to generating capacity. In other words, the more power you produce, the less distortion allowed.
If, then, these non-true sine waves are a problem, why are there inverters that use them?
Answer, it's a classic trade-off. It's technically quite easy to get perfect sine waves from DC. You "chop" the DC to make square waves, then you pass those square waves through large filter capacitors which "smooth out" the square waves into perfect sine waves. There's only one problem. You throw away a large portion of the input energy in the filter. That's a huge waste. Alternatively, you could implement a mechanical inverter using a DC motor to turn an AC generator. This is, in fact, what the earliest DC to AC conversion devices did. Mechanical conversion is also terribly inefficient.
The alternatives (like Trace's SW4024/48 "synthesized sine-wave" inverters) are quite expensive. So, the choices are cheap and "noisy" square waves, or cheap and inefficient sine-waves, or some compromise between wave quality and price. The inverter you choose is going to be one of the most critical choices you make about your system.
![[Image of powerlines]](f_grid.jpg)
Of course you can. Doing so involves meeting the extensive requirements
set down by your particular electric utility. They have very good reasons
of safety and reliability for being very strict in their requirements.
This is not to say that they also wouldn't rather individual producers
would go away, but my expereince has been that people who are open
and involve the utility early and often in their planning get along
quite well with their utlity. Find out who manages grid-intertie
accounts for your utility. Get ahold of that person on the phone months
before you plan to build your system. Develop that relationship. It
will pay off.
In addition to the cooperation of your utility, you will need a "utility interactive" inverter (which the utility might very well have to approve, so don't run out and buy one of these expensive devices before talking to your utility!). You might also require a second meter, a exterior locked cut-off switch, and possibly a few other safety devices. Your utility and their requirements are definitely the place to start!