Solar Powered Site App Notes
Solar powering of equipment is easily done but it is not cheap. Solar powering of a site can easily double the expense of building the site.
Step 1 - Determine Power Consumption
A minimal installation will require one backhaul and one access point. With each drawing approx 8 watts, the total continual power consumption will be in excess of 16 watts(24/7/365). Rounding up is a safe way to approach all solar design.
Round up the power consumption to a "more than" realistic safe level. We will use 20 watts for this example.
Step 2 - Determine Size of Solar Panel
Next decision is how much panel to buy. A bare bare bare minimum of 10 times the continual power consumption is required. Most that have chosen the 10X factor for telecom applications have to go back and add more panels. Only in very sunny locations without snow and frost will this work.
Remember that during the winter in the Northern hemisphere, the higher latitudes have shorter days. Non tracking panels are only producing fully rated power for an hour or two during the middle of the day. On December 21, in Northern Utah, the sun hits the panels at about 9:30 AM and leaves them around 4:30 PM. That leaves 7 hours of some daylight that produces some charging, but when the total energy actually produced is graphed, a half sine wave curve is produced showing very little energy at the start and end of the charging day. Those are the Sine losses which leave you a theoretical max of 70%, which applied to your 7 hours of daylight cuts you down to 4.9 effective hours. This loss happens at all latitudes and all days of the year on non tracking panels. Moreover, low sun angles means the light has to pass through many many more miles of dirty, hazy, foggy air significantly reducing the available energy. That knocks off a minimum of 15% so you are now down to 4 effective hours. Then if you have optimized panel tilt for spring time sun angles, Cosine losses will reduce you another 10% or more. Now you are at 3.8 effective hours. Then consider most days have a few clouds. The is highly climate dependent, but you have to consider the "average insolation" of the site. The NREL has insolation calculators if you really want to get accurate.  Other system losses collude so you end up with about half of the daylight hours worth of charging.
Say you scrimped and went with 10X, 200 watts of panel. During the shortest time of the year, you are generating 200 watts for perhaps an aggregate total of 3 hours (see above, add a few clouds). That is 600 watt hours per day. Your 20 watt load draws 480 watt hours a day (20 x 24 hours). Your battery might be only 80% efficient in delivering stored energy back to you. In abstract, storing your 600 watt hours and then getting them back out of the battery gives you 480 watt hours! Zero margin. One cloudy day and you are out of juice every morning at 3 am. Of course you are not storing all 600 watt hours, some are being made and used concurrently. And you might get more like 85% efficiency on the energy actually stored etc etc. But this rough example illustrates why a 10X system is right on the ragged edge of failing every single morning before sunrise.
If you don't want to be out in the snow with a generator during extended cloudy periods, 20 times is a much more safe figure. Another reason to use the 20X factor is that during overcast days, the panels will still generate at least 5% of their normal power. That means they are pulling the load during daylight hours and not depleting the battery during the day when it is cloudy.
In this example, a 20x factor will be chosen. That yields a 400 watt panel array. Normally solar panels can be found at around $2.50watt or less (3/30/2011). So, $1000 minimum will be spent on the panels.
Step 3 - Determine How Much Energy to Store
Next decision is how much battery to buy. How may days can you go without any solar charging? That depends on climate, weather, latitude, seasonal patterns, topography etc. One week of autonomous operation is a minimum. 20 watts continual draw times 24 hours = 480 watt hours per day. Again, round up to 500 watt hours/day times 7 days is 3500 watt hours you need to store. Storage batteries can be had for as little as 20 cents per watt hour but 30 cents is a mor common figure. So the minimum cost for this battery will be around $1050.
If you have to scrimp, do not scrimp on the battery. A rule of thumb is that half your money will be spent on the panel and half on the battery bank. The 10 X system outlined above will work OK only in very sunny climates with light winter and cloudy seasons.
In the Rocky Mountain region of the U.S. 20 times power and 2 weeks of battery is more of a practical minimum. Even at that, there may be one day a winter where an aux generator may be needed to top off the battery.
Step 4 - Select System Voltage
Note that nothing has been said about voltage. That is because the design and implementation of a PV solar powering system does not depend on voltage. You only need to define the amount of POWER you need to produce and the amount of ENERGY you need to store. Once you have defined that then you simply choose the best voltage for your system. Motorola Canopy products have the lowest power consumption at approx 24 volts. They will operate over a voltage range of about 10 to 32 volts but the lowest consumption comes around 24-26 volts. This is a very convenient voltage for solar systems as most are built in increments of 12 volts. 12,24,36,48 etc. Controllers are widely available for these voltages.
Divide the watt hours you want to store by the system voltage to determine the amp hours of the battery. 3500/24=aH
Telemetry to monitor the voltage of the system is a must. The Packetflux Technologies Site Monitor product does not consume much power and will help you to keep an eye on the system to predict an outage. The only true downside is that an ethernet hub will have to be added to allow the connection of the telemetry board. That will consume a little power.
Step 5 - Select Battery Chemistry
Battery selection is important in that battery technology varies greatly. Any battery with a liquid electrolyte will freeze in the cold weather when discharged. Sometimes a deep discharge is unavoidable. Deep cycle lead acid batts are not very good in the cold weather. Auto batteries will not take the daily nighttime discharge cycles for long. Avoid gel cells totally. EnerSys Genesis batts are designed for operation when frozen.
Flooded batts generally have an operating temperature range of 30F to 130F with greatly reduced life for upper temps. They are generally good for 3500 cycles if only discharged 30% or less. So they are good for 10 years if you can keep them temperature controlled. Unfortunately solar sites don't have the power for air conditioning and heating so generally flooded cells are out. That includes sail boat batts, golf cart batts and car batts.
Ni-Cads are worse for the number of charge/discharge cycles. Typically 1000-2000 cycles at a 20% depth of discharge. They also need to be in a temperature controlled environment.
Valve Regulated Lead Acid (VLRA) Absorbed Glass Mat (AGM) take temperature extremes much better than flooded cells or Ni-Cads. Generally they will work to 5F to 120F. Upper temps are still a problem but they continue to work well when most other technologies are frozen solid. They are not stellar performers as to the number of cycles they can do. Similar to Ni-Cad in that respect.
Unexplored at this time (3/30/2011) are Lithium batteries. Lithium batteries are very expensive, but have a significantly better charge/discharge performance compared to most other rechargeable batteries.
Step 6 - Charge Controller
Choose a high quality charge controller. This will make the most of the energy available and ensure the batteries are as charged as possible whenever there is any solar energy available. Just do a search for "solar charge controller". Charge controllers are generally rated by the amount of current they can control, that means amps or amperage. You want to make sure to get one with enough Amp capacity. If not you may let the smoke out.
This is a good reason for higher system voltages. All other things being equal, perhaps you have the option of powering a system like Canopy off of 12 or 24 volts. The battery will physically be the same size for the same wattage and chemistry. The panel will be the same size for the wattage, it is all in how you wire them up. But, your current will be double on the 12 volt system. That means you will have to buy a charge controller twice as large as the one you would use for the 24 volt system.
Higher system voltages can save you money on the controller and save you energy due to less loss in the wiring resistances. Always go up in voltage if have the choice. You can convert one DC voltage to another if you have multiple voltage loads you need to power.
Step 7 - Talk to your Accountant
Panels age, especially amorphous panels. Monocrystaline solar panels will produce full rated power for more than 10 years. Generally they can be counted on for good service out to 20 years. Polycrystaline solar panels will age a little faster, 15 years is a good rule of thumb. Amorphous start to age as soon as you start to use them. At the 10 year mark they may only be producing 50% of the original value.
Likewise, batteries age fast too. In a non temperature controlled environment, count on changing them every 5 years. Heat is the enemy. If you can keep them cool, you might get a few more years.
In any event, have your accountant expense the whole system if you can. At a bare minimum try to convince them that a 5 year depreciation schedule is realistic. Start a sinking fund to replace the batts in 5 years and the panels in 10 years.
(Spoiler Alert, mild political opinion follows): Contrary to what environmental pundits are wont to claim, PV solar is not a good choice for being green. There are hazardous and toxic components and chemicals used in the production of PV products and batteries. Moreover, they produce a significant waste stream due to their relatively short life cycle. They cannot be recycled (the PV panels that is). Furthermore, dollar for dollar, they produce a pittance of power compared to other technologies. Wind is much more friendlier but is not nearly as reliable and in some areas is wholly impractical for powering a Canopy site.
If money is a huge issue, you can always build a minimal solar system and add a remote starte propane powered generator. That will get you the most bang for the buck. In the winter just keep an eye on the battery voltage. If things are sized right, you can start the generator for a few hours and put in several days worth of charge which should tide you over until the storm passes or you get up there to knock the snow off the panels.
Panel Mounting Notes
You can use a variety of angles to mount the panels. But if they are not on a tracker you almost always want to position them facing due true South. If they are in the shade in the winter for part of a day in the winter due to topography then you would offset them to the West or East to split the difference between shade and and of day. But normally due South is optimal. (Caution: math alert) The tilt angle affects the amount of power they produce. The energy is reduced by a factor equal to the cosine of the angle between the normal to the panel face and the sun.
Don't worry about all that too much because it is not a huge factor. If winter is the problem, make the panels tilt to be at right angles to the sun on Dec 21. That will maximize the power on the shortest day of the year.
You can get this data at the following University of Oregon site: 
To properly shed snow as soon as practical, the panels need to be steep as practical. 45 degrees or steeper for certain. In Northern Utah, on Ded 21st, the sun is about 26 degrees above the horizon. That means that the panels will be at right angles to the sun if they are stood up and tilted back 26 degrees from vertical. Some would call this a 64 degree angle. Pretty steep for sure. But panels that steep will shed snow well. Normally you would optimize them for more of a Feb 20 thus making them not quite so steep but getting a bit more energy out of them the rest of the year. You only lose 1% of your Dec 21 power production if you optimize the angle for Feb 20th. Or you could set the panels at 45 degrees to make the mount easy to build and not so tall. In that case this example would lose 6% of its power capacity on December 21st.
In the summer we have the opposite effect, June 21st has a sun angle in Northern Utah of 74 degrees. That is 29 degrees above a panel tilted at 45 degrees. So they are making 13% less than optimal but long days more than make up for it.
However if you optimized for Dec 21 with the steep 64/26 degree panel, that would make the sun 48 degress off normal on June 21. Cosine losses will be 33%. Still, the long days will help make up for the 1/3rd loss of power due to the steep tilt angle.
If you have not made the effort to get them steep enough, you may have a week of sun in the winter with no useful energy production due to snow accumulation. There has been some work done with remotely activated glycol deice spray systems to remove snow and ice. Keep your eye on this space for details.
Turnkey Solar Power Systems
If you would rather leave the design and building of the solar powering systems to others, several companies are creating Motorola Canopy specific systems. Sunwize (source of photo at head of this page)
Typical Site Power loads:
I am running the following:
- 1x RB532
- 1x RB333
- 1x XR5
- 2x XR2
- 1x TR5a
- 1x RB192
- 1x SuperRMS2
= .3710A @ 12VDC or 4.452W continuous.