Remote Solar Project Review

Solar Powered Parking Lot Security Cameras

Just decided to document my first remote solar powered surveillance system and post it here due to the lack of information (mainly due to the newness of the whole process) here on IPVM.

finished

Overview of Project:

4 poles with a total of 7 cameras all tied back to a main building with wireless radios where the NVR is stored. One pole has 3 cameras, one pole has 2 cameras, and two poles have 1 camera each. The light poles are 277v/480v and surrounded with concrete, brick, and asphalt. The cost to cut/trench all the way to the poles and just run hard cables was about $60,000, versus the $20,000 I quoted for the entire project with solar. The poles are ~6 inches in diameter and steel, so they are very sturdy to mount to. Personally even that quote was kind of high mainly because I wanted to say no without saying no, but they still said yes.

Product line up:

  • 7 SNO-L6013R “Samsung-hanwha-techwin-wisenet-III” cameras
  • 5 Ubiquiti NSM365 wireless radios
  • 2 140watt Tycon power solar kits
  • 2 70 watt Tycon power solar kits
  • 1 16 channel “SamWha technet III” NVR 1637S
  • 2 4TB WD black hard drives.
  • 12v 150watt inverters
  • 4 Ubiquiti tough switches TS-8-PRO

On the solar kits……

So when I initially ordered the solar kits they were supposed to be 12v kits with four 6 volt gel sealed batteries, which is great and fine but they sent me 12 volt batteries…… which is fine as well but when I ordered the kit it first said batteries not included, understandable, but I ordered four 6 volt batteries for each kit to go with them and not only did I get all 12 volt batteries but when the kits arrived they stated they came with two 12 volt each….. so I have 8 extra batteries now. I will say I cannot fully blame Tycon because it could just as easily be the distributors fault in this case as details were lacking.

box1

Product Inspection

All the cameras, NVR, inverters, and radios are things I usually order and they were all fine, but my main concerns were the tough switch and the Solar kits. The weather here in North Carolina while never really in the extremes (less than 0F or higher than 110F) we do however get extreme swings in the weather. What does that mean? it can change ~30F in two hours along with unholy amounts humidity, and this happens all year long. So things out side tend to take a beating rather quickly. The Ubiquiti tough switch however looked like it could take some abuse, but what about the solar kits?

Looking at the weatherproof enclosure, it is rather well built, solid and has lots of mounting rails/tabs for just about anything. Comes with one rail mount for the charge controller (which is included), and room for one or two other things you could mount on the rails. The charge controller comes with one PoE (802.3af) output and one PoE (802.3.af) input for charging. You cannot however use a switching PoE power input, you must use a non-switching (power all the time) power supply. It comes with an additional PoE injector jack. The only difference between the 140 watt and 70 watt model is one extra solar panel, so they are the same with the exception of there is an extra cable, double panel mount and a panel over the 70 watt model. So each kit cost about $1300 for the 70 watt model and $1500 for the 140 watt model, and each 12 volt battery was $265 after that for a total of about $1800-$2000 for each solar kit with four batteries in each enclosure.

Bench Testing the Units

So with the exception of the solar panel, my plan was to test all the rest of the equipment with a 12v power supply (14.8v dc @ 2amp) input as to charge the batteries. But before I would simulate daily use under normal conditions, I decided to hook up 2 batteries, 2 cameras, one radio and one tough switch and see how long it would run till the batteries would die. The results exceeded my expectations, as the charge controller cut off after 47 hours of use. This is of course with just two of the batteries instead of the four I had planned to use, which mean I can easily get the 3 days I was aiming for. I put the units on bench power to charge overnight and started working out how I was going to mount them to the poles. On a side note, I choose the SNO-L6013R model cameras due to the low power usage (5-6 watts) versus some of the higher end cameras that tend to pull 10-14 watts. Having installed solar systems far before I ever started installing anything security related helped out tremendously in designing and picking out materials for this project.

Installation

So having all details of how it is going to work hammered out it is on to the install. The pole is 6 inches in diameter, my kit supplied straps that were 5 inches max diameter. Luckily they gave me 6 straps so I could double up the straps 3 times. The solar panel came with just two 4 inch straps but I had extra ones laying around so not a big deal.

I got the box up:

box

Then the solar panel

panel 1

Then the camera

panel and camera

Then the radio

radio

Then the batteries

guts

This beast is heavy so I am going to put a bolt though the pole and bracket at the top and bottom just to be safe. Made all my connections, four 12 volt batteries in parallel, and everything is up and running. Tie cameras into the NVR and there is almost no lag or broken frames what so ever, I will update in a week and a month with any issues I have.

Final Thoughts for the Moment

I have to say I am rather impressed with the Tycon power units, I wish the documentation was better, but that could be more on the distributor than Tycon. All the other products were things I have/had used before with the exception of the Ubiquiti tough switch, but I have used enough of their products to know that it was more or less going to be what I needed. Not sure about Samsung-Hanwha-techwin-wisenet future’s is going to be but I do like using some of their products in cases like this.


Great detail, Eddie!

One question: I see why you opted not to trench, but why not just get the power from the lights, using a 277V light pole tap; and either power the system (if the lights always have power) or charge the batteries (if the lights are switched)?

Again, nice report!

all the lights are fired from a 50 amp lighting contactor inside the main building where a timer is connected. Timer is only on at night hours, I did have a quote to install transformers and outlets to charge the batteries at night. but its slow going with the power company to comply. I was fortunate enough to get permission to use the poles so I was only going to be but so greedy.

No worries.

Do the solar panels charge the batteries at night from the lights at all? A little extra guaranteed power can't hurt.

It looks like the summer could bring a bit of high foliage to the area; what did you use to estimate the effect of that?

yes I have in formed them they have to trim the trees but we wanted to wait till they are mounted to make sure we trim what we need and not deforest ( its a botanical garden)

I have not checked if the panels generate power form the lights yet, I am not due for some second shift work for a while but I will keep in mind if i can

That's awesome! Thanks for sharing.

Related: Solar Surveillance Guide

Cool! This is neat! Those Tycon enclosures look nice. A couple of questions:

1) Do the enclosures lock? It looks like they might have some kind of barrel lock.

2) Is the inside of the cabinet heated?

3) Do you hang lightning arrestors/ or ground this any certain way?

4) How is video pulled from the NVR?

1) yes they do, they come with a sturdy barrel key that you should hope to never lose

2) no but everything int the cabinet is rated -30C to +60C or -22F to 140F except maybe the invertor but you could get one to do that if you wanted or install a heater I suppose but you would have to drop 2 batteries out of the unit to do it.

http://tyconsystems.com/index.php/17w-continuous/344-rpst1212-100-70

http://tyconsystems.com/index.php/35w-continuous/336-rpst1224-100-140

3) the pole is grounded to the rebar in the concrete pillar it is mounted to, and I put 2 bolts though the mounting bracket and everything inside the cabinet is grounded. There's also a surge/lighting protector before the hardened switch which powers all the cameras and radio.

4) right now they use the NVR as a monitor so with the mouse mostly. I will add I have SD card in all of the cameras on 24 HR record in case a radio goes out

Eddie, what's the purpose of the power inverters?

Since the panels are generating power at around 12v, wouldn't the most efficient thing to do be just run everything at 12?

By using cameras that can also run off 12V? And a non-POE switch that runs off of 12v?

Otherwise isn't there a bit off loss in converting everything up to 110v just to go thru a wall wart and come back down to 48v at the switch and ultimately down to 12v at each device?

yes there is always loss when you convert power and i did look for 12v cameras but I ran into other problems with configuring in all the other parts i needed. I knew it was not going to be the most efficient build but the was due to lack of available parts that would work for me. ill answer this more in detail below

I can not see how the numbers match up on this. In this region, the average daily sun is about 4 hours. If we look at the site with 2 cameras and 140 watt panel, and assuming that you can get best case get the 140 watts, that is 560 watt-hours per day. Cameras (12 watts?), Radio (5 watts?), and switch (6 watts?) as well as POE/Inverter losses (etc - 7 watts) may bring the prospective load at 30 watts (24/7). This exceeds your solar power capability. This does not bring into play the 3+ days of "no-sun" periods (cloudy/rain) and the need to keep batteries above 50% charge. I am not sure what the amp-hours of these batteries are, but I have found that even a single 100 amp-hour battery in Maine with single camera/radio solution (MPPT controller) is not enough on multiple "no-sun" days.

I think one of the big problems that can be seen with this is the inefficiencies with the components used. Some ways to remedy this:

  • Power cameras from 12 vdc instead of POE. (Turn off IR if used).
  • Power the radios from 12 vdc (I assume they are powered by the Tough switch)
  • Get rid of the Toughswitch. I am not sure of excess power draw, but with the OS and I am assuming you are going 12vdc - 120 vac - 24 vdc to power this, has to be quite a bit of unneeded power (and losses)
  • Use dual NIC ports on NSM radio versus switch. (On 3 camera site, use ultra low power 12vdc switch or even a second radio)
  • Use an efficient MPPT charge controller

By doing this you may shave this down to (6+6+5+3 = 20 watts) 30-40% less continuous draw, which requires 480 watt-hours per day. You would probably still need 280 watts of panel plus 40 amp-hours per day of battery (consider 3 days no sun run time plus discharge to 50% battery plus battery inefficiencies in cold weather - 250-300 amp hours - or three full size 12 vdc-100 amp hour gel cels).

I can not see the 70 watt sites running even a day (Nov-Mar). If you are installing this during the longer days (sun from 7am - 7 pm) versus the shorter days (sun from 9am - 3 pm), get the sign-off quick. In December/January, where it is cold (bad battery efficiency) and with minimal sun hours, this system will be down more than it is up. I assume you will not be tilting these based on the seasons, which makes matters even worse.

Once other question, though. Are the 3.65 GHz radios properly licensed? Why not use NSM5 radios?

your numbers are a little of from i tested on my bench with the equipment I had the cameras I used pulled 5.6 watts or 6 watts each at 802.3af and on highest setting of 1080p, tough switch pulled 3 watts with nothing in it but turned on and the radio pulled 5 watts. the 70 watt models are only getting one camera attached to them. so the math is 6 + 3 + 5 = 14 watts which allows me 3 watts wroth of loss to occur as the 70 watt panels are rated for 17 watts of continuous power to stay in the range of being able to charge back what is used on a daily basis. the 140 watt models are rated at 35 watts of daily usage.

As I said before I this running on my bench with half the batteries loaded in to see just how bad the losses were going to be. which is why I got 47 hours with 2 cams, switch, radio, and the invertor before the charge controller shut off. It gave me a really good feel for how much I was using and how long I had between any kind time before it had to charge the batteries.

Originally I wanted to go the full 12v route but at the time I had trouble finding what I needed to accomplish this. 12 v cameras are not a big deal to find but they do pull on average about 10 watts. radios on the other hand are not so easy to find especially when i have local Regulations that say I cant use the 2.4GHz and 5GHz bands but fortunately I work for the government and while I have to register that I have licensed radios I dont dont have to pay all those nasty fees that come with them. So more or less I am stuck with the reliable Ubiquiti M365 series unless I want to suffer with 900MHz radios again.

at the time I ordered this equipment they had not released some of the nice stuff that out now like this:

http://tyconsystems.com/index.php/5-port-passive-gigabit/530-tp-sw5g-nc

yeah that runs of 12vdc would have been really nice to have.

I dont if you live in the north pole or what but the average daily full sun i have here in NC, USA is 6 hours winter time and 12hrs- summer

batteries are 12v 100ah each wired in parallel for 400ah total or about 2400 watts before the charge controller shuts off. on the cold weather thing yes i expect to see some problems but after inspecting quality of materials used Unless it get to the lower end of the manufacturers threshold I cannot see any issues arising besides lack of sun but as I said earlier I am going to update this in a week and a month from now.

they also have a bunch of new stuff now like remote power monitoring devices as well that I may check out.

at the time I ordered this equipment they had not released some of the nice stuff that out now like this:

http://tyconsystems.com/index.php/5-port-passive-gigabit/530-tp-sw5g-nc

yeah that runs of 12vdc would have been really nice to have

Eddie, I'm not sure how would that switch help you. If you power it from the battery it's just going to output 12V Passive POE.

In any event you already have a two port wired switch the radio, I believe. If it's in bridge mode, it should be able to connect the 1 and 2 camera poles without need for an external switch.

Yeah that one wouldn't for what ever reason the right page didn't load might have been my fault hopefully the spec sheet will load.

http://tyconsystems.com/documentation/Spec%20Sheets/TP-SW5%20POE%20Switch%20Spec%20Sheet.pdf

its the TP-SW5G-24 model but i would probably wire the batteries in a 24 volt configuration though just to make easier to grab power for the Ubiquiti radios that run on 24vdc

This perked my interest. I am looking back at a project that I put in 2.5 years ago. Recently, the 100 amp-hr battery was spent and had to be replaced, but, apparently, the camera has only been down a couple of times. I am assuming that the 100 amp-hour battery was not enough (I calculated twice this) and was reaching < 50% charge, which is why it probably wore out. Keeping a battery about 50% charge will prolong life quite a bit. (I actually ordered two batteries for project, but used one, just to find out - guess I found out the answer).

Project was:

  • Morningstar MPPT 15 amp controller
  • 140 watt Kyocera Solar Panel 12 vdc
  • (1) Vivotek 5 MP bullet IP camera (has to be bullet so no heaters needed, IR turned off, powered from 12vdc on pigtail)
  • NSM-5 radio (run off 12 vdc from batteries)
  • Site is Brunswick,Maine - panels tilted at winter angle - horizon not obstructed by trees - in open parking lot (probably sun restricted 2 hours before/after sunset east/west)

So you can see, this barely worked (some outage days) and definitely not enough battery. Of course this is Maine, and there is probably some snow coverage on the panels in the winter. Customer reports couple of outages outside of this season, probably during a couple of days of clouds and snow. (In Maine, normally 5/15 - 6/15 is very cloudy). Solar comes through clouds in some fashion.

I think in Eddy's case above, using a charger and transformer from the light pole would probably be a more effective solution. The rectifiers/inverters/switch are inefficiencies that will probably cause problems. I have tested 70 watt panels and outputs in the summer (Maine). Even then, 6-7 hours of power above 50% of the panel rating was all I could get (July) and only a couple of hours middle of day approaching the panel rating. I was able to test and run a 70 watt/camera/radio solution with 100 amp hour battery/MPPT controller in the summer, but it would not work fall/winter/spring (battery wore out after several months).

It looks like the unit in the picture is surrounded by hardwood trees, which have not yet foliated. Do they exist in the direction the panel is facing? In NC you will be at 38 degrees in the winter and 78 degrees peak of summer. Is there clearance for a full day of sun?

Just one battery? yeah that would have been your first problem. when ever you size up power systems it is no more than 80% from the power company or use no more than 40% from back up sources, battery and solar are considered to be back up.

then next project I do i am going to try and use some of the new equipment and make it as efficient as possible with power monitoring. I have have a feeling now the the word has gotten out locally that I have installed working solar security systems their will be more request for them.. unfortunately

Even then, 6-7 hours of power above 50% of the panel rating was all I could get (July) and only a couple of hours middle of day approaching the panel rating.

You weren't charging your batteries at night also, though. ;)

while I am thinking about it the 17 watts daily usage is rated based on having at least 4-6 hours of full sun a day 3-4 days a week so that you dont use more than you can put back. Obviously the weather isnt perfect and you can use more than 17 watts a day but this is the recommend math when they claim you can have 17, 20, 35, etc of daily continuous usage.

Eddie - this is an excellent write up that should help many designers. We've done a number of solar installations with cameras and LPR's. We haven't used Tycon yet, choosing to build our own enclosures. However, after seeing what you've done we will take a closer look at them. Thanks for sharing!!

Eddie, I am wondering how you went about insulating the battery terminals? Seeing the din mounted devices on the inside door panel makes me nervous. You guys don't have any earthquakes up there do you?

no earthquakes but the cabinets are really versatile and spacious the orange board you see in the back is spaced off the enclosure about an inch. so if you push all the batteries to the back of it you have about 4-5 inches of space before the door. both the devices are plastic so not worried about that. you could if you wanted secure the batteries down if you wanted to.

The nicest thing about whole kit that impressed was the cabinet, its a really nice cabinet lots of options you can do with it.

I've seen two group 31 AGM batteries dead shorted and 10,000 amps of arching isn't very pretty. It was quite scary.

i have melted a battery terminal or two its not fun but the worst i have ever seen was in a basement after some one crossed the battery backup of about 300 AGM batteries to the main building power supply. there were beads of copper all over the floor and the cinder block walls were copper plated with a silhouette of a human figure in the middle of it. that kinda thing, it gives you a healthy respect for batteries and back up power connections after that.

no he did not make it.

Impressive Eddie and thanks. I do have some questions. The poles look like Duke Power utitity poles? Or are they owned by the Gardens? Did you get permission from the Utility or do they know about this? In our experience, Duke Power can be very difficult to work with. Unless the projects are public, they tend to say no.

Another question is about the foliage. There are no leaves on the trees in your pictures. How do you think the power, radio distance and camera pictures will be effected when May gets here and the trees are full of leaves and blooms of all kinds?

Again, thanks again and I look forward to your answers.

they are and they are not. We ( the state) buy those poles from the same people duke power does or from duke/progress themselves. as long as they are on state property they are our to maintain and we have people to do so. we try to keep as close to the guide lines the duke power asks as you can tell they are marked similar, but at the end of the day they are our poles and we do what we want with them.

I would say you may have to put you own poles in most cases.

I have not had any problems with the m365 series, but I have had interference with the 2.4GHZ models. I have put one under some wooden steps and have zero signal loss, so not really but I will update in month when all the stuff starts blooming and the panel is covered in pollen.

I dont if you live in the north pole or what but the average daily full sun i have here in NC, USA is 6 hours winter time and 12hrs- summer

I looked at various sources for average winter solar hours for design (NC), and all come up with 4 to 4.5 hours of sun (I am assuming that these numbers correspond to the correct tilt angle). Design has to be based on worse case average (winter). Even if you are only using 17 watts of power (should test with ammeter from battery load), that means you need 400 watt-hour or 34 amp-hours per day. The 70 watt panel will provide on average 315 watt-hour or 26 amp-hours per day (4.5 hours), and that is unobstructed E and W horizons (I mentioned the trees visible in the pictures above). This is assuming that the 70 watt panel are operating at 100% efficiency.

So right out of the gate, in January, you need 34 A-hr of power per day and you only have 26 a-hr. Add to this a couple of days of cloudy "no-sun" days in reserve (there is some solar on cloudy days, but at a fraction of full sun days) to keep batteries above 50% of discharge. For multiple cloudy days (lets say 3 days with solar at 20% output), you would need almost twice the battery capacity (for the one camera solution) and solar capacity of 34 A-hr plus the excess needed to bring the batteries back up quickly in maybe 1-3 days to prepare for another 1-3 day outage.

Of course there is only so far you can go with this if you get long durations of cloudy weather (April-May), but at least you will not be replacing batteries on a yearly bases. You can "overdesign" this, for sure, so you have to be practical and expect some outages during the year.

Removing the unneeded components (voltage converters, switches) would drop the demand quite a bit.

Actually, the Tycon calculator is not too shabby (will have to favorite this site). Entered the below for 4.5 hours sun, from their location look up, 4.0 to 4.7 hours, and 2 days battery backup. It is curious that it does not account for the power required for their equipment (charge controller/voltage converters).

http://tyconsystems.com/index.php/resources/tools

Average Power Consumption: 17.00 Watts (W)
Daily Power Consumption: 0.408 Kilowatt Hours (kWh)
Minimum Size Solar Panel Needed: 91 Watts (W)
Suggested Minimum Battery Capacity: 12V 192 Amp Hours (Ah)*.

You would have to live in AZ or NM to get 6 hours of sun (very few clouds, trees, and elevations down there). Tycon is being kind of sneaky with the specs, basing the 17 watts continuous on 6 hours of sun, and the 35 hour backup @ 17 watts on completely draining a 102 amp-hour brand new battery.

With the 47 hour test, it does show you are using about 24 watts on that setup (2 cameras, switch, inverter, etc) if the controller will run down to the end of the 100 amp-hour life. You made no indication if video was streaming and the radio was active for this (maximum achievable bandwidth). I am assuming that the cameras were tested with no IR and that the power ratings for the camera (5.7 watts) are for IR on, which, if so, would indicate higher power usage for the controller, switch, and inverter.

Seems like kind of an expensive project to be utilizing $130 1/3" fixed lens cameras.

Eddie, you seem genuinely open to suggestions, (which is rare).

If I were in your shoes at this point, I would consider adding the small expense of the LPT to grab the 277 off the tap and use it to charge the batteries in addition to the solar panels.

They're about $50. The batteries would only have to last for a day this way and the solar would provide redundancy in case of power problems.

Good luck and thanks for the all the info whatever you end up doing!