Solar setup for forestry research

bassman · October 2013

Guys- forgive me if I do am doing any calculations wrong here or dont understand your responses. I am not well versed in power setups.

I have a off-the-grid solar setup in a Florida planted pine plantation where we are conducting some forestry research. Solar panels are mounted on a 15 foot tower (trees are 30-40 ft tall). I have cleared several trees around the tower and pruned back other trees to attempt to receive as much sunlight as possible. We are having trouble maintaining power with our current battery setup and I am seeking advice for a new battery bank.

I have attempted to estimate my draw, and then, based on the setup I have in place, decide on what type and quantity of batteries I need.

Some notes:
1) The draw is fixed. I draw the same amount of Ah 24 hours a day, 7 days a week.
2) The type and quantity of solar panels is fixed. They are mounted on towers in the air and I cannot add any more

With that, it appears the only major component to the system I can change are my batteries???

My system:
-I am using (3) BP solar panels with Pmax of 7.1 A, 125 W
-I am currently using (2) Everstart flooded deep cycle batteries, 12 V 100 Ah, wired in parallel
-I have Morningstar ProStar 30A PWM solar controller

I used a worksheet to get some of the numbers below:
-Based on the sensors, power boards, dataloggers, etc collecting data, I estimated 2.613576667 Amps/hr draw
-I scaled that to 62.72584 Ah/day, and with a battery loss factor of 20%, I got 75.271008 Ah/day
-I choose a random 3.5 dark days we want to support (dont know how to choose this number, as we never have days that are truly 100% dark, though we do get some consecutive overcast days...I assume you get some charge on these days?), thus we need 263.448528 Ah to ride through those 3.5 days
-I used a depth of discharge of 50% and corrected my Ah to 526.897056 Ah based on the DoD.
-If used 100 Ah batteries, my worksheet estimates I need 6 batteries

The flooded batteries we are using now are heading towards 2 years old have been abused pretty hard. Sometimes I cant make it to the site for a number of weeks, and thus they get drawn down pretty low (10-11 V). Last summer the setup ran well and took little maintenance in terms of swapping out batteries. The winter wasnt too bad, but required regular swapping of batteries (cant recall the interval). Currently this fall we are having to go to the field every 2 weeks or so to swap out batteries to keep the system running, and this is surely to get worse with winter. I think with the system we were running we severally shortened the life cycle of our batteries, despite attempting to swap them when battery voltage got low and adding DI water to them.

All that said, I have a few questions:
1)
a) Is possible based on my fixed solar panel setup of (3) 7.1 A 125 W panels, fixed solar controller, and fixed draw of 2.613576667 amps/hr, for me to calculate how many batteries I would need to never have to go to the field to change batteries, using a battery DoD of 50%? By my calculations I know I can determine how many batteries I would need if I had no panels to power for X days, but I am not sure how to factor in the power the panels provide. One factor to consider, in the winter with the lower sun angle, and because my panels are under the canopy, sometimes they only get patchy sun.
b) How about if I not worry about DoD, but instead just never want my battery voltage to drop below 11.8 V (what my sensors need to collect accurate data)? ...I am sure not how 11.8 V would correlate to DoD.
2) We have looked into switching to AGM deep cycle batteries and ditching the flooded deep cycle batteries. Several other groups that do our line of work have ditched flooded batteries a long time ago. I have read into the pros and cons of flooded v. AGM v. gel. Is there a reason why AGM batteries are better for our scenario where we are constantly drawing power and sometimes being abusive and letting the battery voltage get low? AGM batteries seem nice but also appear quite expensive. I dont want to under-engineer something and drastically shorten their life cycle
3) Does anyone have a recommended brand of AGMs. It appears their price varies from Crown to Deka to Trojan to Werker, etc. I also seed they come in flat-plate and tubular forms
4) Any recs on where to buy AGMs online?

jcheil · October 2013

Re: Solar setup for forestry research

Curious, are you saying that you run the sensors and such directly from the 12v batteries?

Just a thought: I know there will always be losses adding an inverter, but using a real small one, then powering a 110v->12v regulated 12v power supply to power the devices would give you constant 12.0 volts regardless of your battery voltage all the way down to 10.x volts when the inverter shuts off due to low battery voltage. It will also prevent the sensors from seeing any more than 12v, which could be as high as 14.x+ volts based on the controller's settings (which might damage the sensors also). Now just for the record, you should never let the batteries get to the point where the inverter shuts off. I was just stating that as an example.

And I wonder if those batteries are REAL deep cycle batteries or one of those marine-type hybrid batteries that are not really good for solar at all and would be lucky to last 2 years. (2) 6v trojan T105's would give you 225ah of storage. Someone else can do the match as far as if you have enough panels, etc. Just my thoughts.

bassman · October 2013

Re: Solar setup for forestry research

Sensors are not directly connect to the batteries.

Battery 1 to Battery 2 in parallel, ran to a datalogger. Datalogger has ports that branch to two multiplexers. Multiplexers have powerboards wired into them, powerboards have sapflow sensors wired into them. The dataloggers, radios, multiplexers, etc have very little draw. The powerboards, which heat a thermocouple in the sapflow probe, are responsible for 99%+ of the draw.

I have not considered using an inverter. As far as I know, our folks have never used one.

The batteries are indeed cheap Everstart batteries from Walmart. Labeled as marine deep cycles.

jcheil · October 2013

Re: Solar setup for forestry research

bassman wrote: »

Sensors are not directly connect to the batteries.

Battery 1 to Battery 2 in parallel, ran to a datalogger.
The batteries are indeed cheap Everstart batteries from Walmart. Labeled as marine deep cycles.

Sorry, that is what I meant, I used the wrong word; meaning the equipment, not the sensors are all powered directly from the batteries. And yeah those batteries are not designed to deep cycle as many times as needed for solar use.

Again, one of the experts please chime in as far as the calculations as to enough panels, etc.

BB. · October 2013

Re: Solar setup for forestry research

On my Phone at the moment... Any way you can run the thermistors for five minutes once an hour (as an example).

You probably do not need minute by minute data.

-Bill

bassman · October 2013

Re: Solar setup for forestry research

Powerboard have to provide constant heat to the thermocouple in the tree. This cant be altered unfortunately. This is my big draw.

The 1 min data collection interval is trivial. I could collect every 10 seconds if I wanted, as the active and quiescent current drains from the datalogger, radio, mux, etc is very small (on the level of 0.00X to 0.000X Amps per hour for each piece of equipment).

Thanks for the responses so far.

mike95490 · October 2013

Re: Solar setup for forestry research

you could change the PWM controller to a MPPT controller, and maybe up your solar harvest 20%.

And I like the suggestion to use 2, 6V batteries in series, to get 12V @ 220Ah. May be a bit more robust than the marine batteries.

But you really need more solar PV.

niel · October 2013

Re: Solar setup for forestry research

i'm going to agree with mike on this as those not dark days are far down in power production and eventually when the sun returns there's insufficient power to recharge the batteries and run the loads.

for the approximate same area footprint your present pvs take, you could place up to 2 of the 60 celled variety of pvs with up to 270w each. this into an mppt cc could significantly up the amount of power available for your loads and batteries.

also, check for shading problems in the colder months when the sun is lower in the sky as this could rob you of enough to keep your batteries up too.

zoneblue · October 2013

Re: Solar setup for forestry research

bassman wrote: »

-Based on the sensors, power boards, dataloggers, etc collecting data, I estimated 2.613576667 Amps/hr draw
-I scaled that to 62.72584 Ah/day, and with a battery loss factor of 20%, I got 75.271008 Ah/day

Lets forget the batterys for the minute. If you dont have the power source, no amount of battery is going to run the gear.

So lets get this straight. Your load is a not negotiable 75Ah/day?, which is 900Wh at 12v nominal. Assuming you have for the sake of the argument, 3 sun hours, and yours may be less, a good deal less, by the sounds of your shading. Then 3 times 125Wp times 3 hours is 1100Wh or thereabouts. Thats the average production figure, without accounting for any losses yet, thats cutting it too fine. Id say way too fine.

You dont say what the constraint on pv is, but it is clear you need a bit more. Some analysis on the shading will tell you how much more. Id be concentrating on the peak period 10-2pm, asumming thats when your sun shines. If you can get that unrestricted, and your array is sloped optimally thats a very good place to start.

To give you a ball park my system produces 3000Wh on all but 3 days of the year. Scaling that down to your loads, that would be equiv in your case to 675W PV, and 300Ah @12v storage. But remember thats still an optimally sited array, zero shading. And, we are at 39deg lat, so winter sun is not bad. And... you still have those three problem days.

If none of those apply add more of everything. For all i know you are in the canadian whop whops.

Solar setup for forestry research

Comments

Categories