Too low voltage despite plenty of sun / Autonomous weather station

Started by canuck webmaster, June 22, 2017, 01:41:57 PM

Previous topic - Next topic

canuck webmaster

I have an autonomous weather station with solar panel and some sensors and cameras, on top of a hill in Southern Quebec. Current setup is:

- 2 x 140W solar panels, installed facing south, completely vertical
   (at latitude 45° in Canada, this is much efficient in winter: snow doesn't stick)

- 8 x 6V/220Ah AGM battery bank wired in 12V
   - Battery efficiency: 96%
   - Battery temperature compensation reference: 25° C
   - Percentage of change per degree C: 1%

- I have a Midnite Kid with an old firmware (Kid bought February 2015, firmware never updated)
- temp sensor installed, but no more WBJR
- load is around 1A, with occasional peaks at 2.5A when a small heating device is on

The station is operational since July 2016. Everything was fine last summer, and in winter we did experience quite a few very cold and cloudy days, but voltage was adequate (heater is turned off below 12.2V, and below 11.8V everything is turned off, except the minimum (main weather sensor, cell modem, data logger). This setup worked well, no power shortage at all during last winter.

Last summer, max voltage was around 14.4V (during bulkMPPT and absorb mode), in winter it was lower of course, as low as 11.77V on some occasions, which I admit is below 'safe' level for battery longevity. In winter, max 'absorb' voltage on clear sunny days never fell under 13V.

But since mid-April, with long sunny days back and warm temperatures, I noticed the max battery voltage is well below 12.5V (and yesterday, the summer solstice, max voltage was 12.44V at 1pm).

Here is a min/max voltage graph since July 2016:



Here is solar radiation graph:



So there is definitely something strange. Since station is 70 miles away, I cannot go there often to check hardware.

Last time I went up there a month ago, everything looked fine. Nothing wrong with panels, all very clean, cables are fine, battery box with Kid inside were fine too, all components looked like new. No trace of lightning, vandalism, flooding, or animal activity.

At around 9:40am on May 20 (cloudy day) the Kid was showing :
   3.7A BulkMPPT, 12.5V solar 45W
   Input 34.9V, battery 12.5V
   KwH 0.1, Ahr 9.0

So what is going wrong ?
- are my batteries getting too old? They are 2 years 4 month old, AGM are supposed to last 3 to 5 years in normal conditions (are my conditions normal?).
- is one of my 8 batteries going bad ?
- is the Kid battery temp sensor acting funny ? Temperature is now around +20°C (68°F).
- is there some variables in the Kid I should change asap, like battery efficiency?
- what else to check on my next visit up there? I'll bring a voltmeter this time...

Vic

A few quick comments;

You have a relatively large battery bank  --  about 880 Ah.   With PVs  adjusted to take almost no advantage to the solar irradiance that is available on sunny days  (about 90 degrees -- vertical) in the Summer, you are probably getting little production during peak sun hours.   Also the PV array is fairly small for the size of the battery bank.  A maximum charge Rate for Lead Acid batteries is about 10% of 20 hour Capacity.   This would (ideally)  be around 80 - 90 Amps.   There can be some reduction in this Rate on systems that have low peak power demands.

IMO,  you should consider adjusting the PV elevation angles in the Fall and Spring to accommodate snow,   and then the good sun conditions Spring through Fall.

Your batteries may have been damaged.   Parallel strings of batteries can be difficult to manage,   especially for sites that are not well-attended.

Some AGM batteries need to Limit the range of Temperature Compensation,  and some have limits on charging at cold temps.

Many AGMs like extended Absorb times,   to achieve a true full-charge ...   and so on.

A new set of batteries will forgive a number of issues in the first months,  perhaps up to a year,   but often they will scream for attention after some time.   Would suggest that you adjust the PV elevation angle to something that is close to your
Latitude,   or Latitude minus something like Latitude minus 10 - 15 degrees for now,  and try to get the bank charge up.

Parallel battery strings often do not share charge and discharge currents well,  resulting in low capacity,   and the chance (as you noted)  of one or two abused batteries dragging down a/the good ones.
FWIW,   Vic
Off Grid - Sys 1: 2ea SW+ 5548, Surrette 4KS25 1280 AH, 5.25 KW PV, Classic 150,WB, Beta Barcelona, Beta KID
Sys 2: SW+ 5548s, 4KS25s, 5.88 KW PV, 2 ea. Classic 150, WB, HB CC-needs remote Monitoring/Control, site=remote.
 MN Bkrs/Bxs/Combiners. Thanks MN for Great Products/Svc/Support&This Forum!!

canuck webmaster

Thank you Vic for very complete answer.

Yes I maybe have a bad battery or 2. Will check that on site soon with voltmeter. I hope I still have enough good batteries so I can switch to a 6-bank 12V setup...

Re-orienting PV is not going to be easy... But it's definitely a good advice. Thanks.

ClassicCrazy

Quote from: canuck webmaster on June 22, 2017, 03:28:25 PM
Thank you Vic for very complete answer.

Yes I maybe have a bad battery or 2. Will check that on site soon with voltmeter. I hope I still have enough good batteries so I can switch to a 6-bank 12V setup...

Re-orienting PV is not going to be easy... But it's definitely a good advice. Thanks.

For testing the batteries a  load test while monitoring the voltage will be better. I have seen AGM batteries show fine voltages until a load is on them and then the voltage would dive.

What is the temperature compensation value you have set  for them ? It would be something like -3 mv per cell .  The numbers you stated are for the SOC value which is of no use if  you don't have the Whizbang on it. Those are not the temperature compensation value of the charging .  The place where you enter the temperature compensation value also has the temperature limits that Vic was mentioning that you can set if you need to.

Larry
system 1
Classic 150 , 5s3p  Kyocera 135watt , 12s Soneil 2v 540amp lead crystal for 24v pack , Outback 3524 inverter
system 2
 5s 135w Kyocero , 3s3p 270w Kyocera  to Classic 150 ,   8s Kyocera 225w to Hawkes Bay Jakiper 48v 15kwh LiFePO4 , Outback VFX 3648 inverter
system 3
KID / Brat portable

Westbranch

My first thought  since you are generally low on PV input overall is to add 2 more panels that are 'summer oriented' then you don't have to worry if the snow covers them as the others are vertical..  this will insure a good charge over the summer..

PV is CHEAP right  now, especially if you bought your first PV more than 3 yrs ago!
KID FW1811 560W >C&D 24V 900Ah AGM
CL150 29032 FW V.2126-NW2097-GP2133 175A E-Panel WBjr, 3Px4s 140W > 24V 900Ah AGM,
2 Cisco WRT54GL i/c DD-WRT Rtr, NetGr DS104Hub
Cotek ST1500 Inv  want a 24V  ROSIE Inverter
OmniCharge3024  Eu1/2/3000iGens
West Chilcotin 1680+W to come

canuck webmaster

Thank you guys for feedback.

I'm afraid I don't have the "temperature compensation value" in my notes. I'll check that later next month in the Kid when visiting the station on site (2 hours drive, 2 hours hike uphill, one way... I don't do this often!).

Adding some more PVs is not quite an option, PV are cheap, but helicopter ride is not!

I'm considering reorienting my 2 PVs instead. But I cannot change elevation angle, only rotation angle... So I might turn them a bit more to the East, to get more light in the morning, when sun is low. And in Autumn, I'll turn them back facing South.

If my expenses were not limited, I'll go for something like this ("Arpont" mountain refuge in French Alps, photo courtesy DualSun.fr):



But my setup and budget is much more basic...


Westbranch

Canuck Webmaster, you left off :
"2 people (Sherpas?) with Pioneer packboards and 2 - 40# PVs plus a lot of sweat!  :o   ;)

Do you have any pics of setup?

I just did some mental overlaying of the W/m2 to the input and it really looks like the problem is when you fall below 300W/m2 ~ mid October, near the 'cross-quarter day' to ~ mid March, the equinox... the potential solution is the same as before.... more PV but at the winter setting so that you can harvest enough WATTS from that weak sunlight.  Also addition of a shunt and WBjr would go a long way in helping you to get the 'real' charge going into the battery vs just an estimate
hth
KID FW1811 560W >C&D 24V 900Ah AGM
CL150 29032 FW V.2126-NW2097-GP2133 175A E-Panel WBjr, 3Px4s 140W > 24V 900Ah AGM,
2 Cisco WRT54GL i/c DD-WRT Rtr, NetGr DS104Hub
Cotek ST1500 Inv  want a 24V  ROSIE Inverter
OmniCharge3024  Eu1/2/3000iGens
West Chilcotin 1680+W to come

ClassicCrazy

How is your station transmitting out the data ?
Are you a ham radio operator ? You could send telemetry data out to APRS network about voltage and other stuff .

Larry
system 1
Classic 150 , 5s3p  Kyocera 135watt , 12s Soneil 2v 540amp lead crystal for 24v pack , Outback 3524 inverter
system 2
 5s 135w Kyocero , 3s3p 270w Kyocera  to Classic 150 ,   8s Kyocera 225w to Hawkes Bay Jakiper 48v 15kwh LiFePO4 , Outback VFX 3648 inverter
system 3
KID / Brat portable

canuck webmaster

My setup looks like this. There's two 'tripods', one with instruments, one with two 140W PVs. The battery box is in between, on the ground. Weather data (incl. voltage) and images are sent to the nearest cell tower with a 4G modem and yagi antenna:



Around noon in February 14:



March 30:



June 26:



Elevation is 700 m (2300 ft). Lowest temperature (December 16, 2016) was -28.5C (-19.3F) , highest so far (June 11, 2017) +27.2C (+81F)...
Max average wind (March 9, 2017) was 77 km/h (48 mph)... Quite a challenge... Will post more data next month... Stay tuned...



canuck webmaster

Quote from: Westbranch on June 28, 2017, 06:32:16 PM"2 people (Sherpas?) with Pioneer packboards and 2 - 40# PVs plus a lot of sweat!  :o   ;)

Well yes I guess it can be done... with a bit of dedication... 



Mt. Everest sherpa photo courtesy of maclynmilsark.blogspot.com


Westbranch

How have you handled lightning? 

Looks like good attractants for strikes being the highest METAL on the ridge... :o
KID FW1811 560W >C&D 24V 900Ah AGM
CL150 29032 FW V.2126-NW2097-GP2133 175A E-Panel WBjr, 3Px4s 140W > 24V 900Ah AGM,
2 Cisco WRT54GL i/c DD-WRT Rtr, NetGr DS104Hub
Cotek ST1500 Inv  want a 24V  ROSIE Inverter
OmniCharge3024  Eu1/2/3000iGens
West Chilcotin 1680+W to come

canuck webmaster

Quote from: Westbranch on June 29, 2017, 04:36:43 PMHow have you handled lightning?

We thought about it a lot, and we believe a lightning rod is actually attracting lightning, so we didn't install any. But the station has quite a low profile, there's plenty of higher trees around. We took a chance. So far no issue.

canuck webmaster

#12
So this week we've been to our weather station site, and we could check the voltage. Each one of the eight 6V batteries showed normal values around 6.1V~6.2V. No battery looked damaged, maybe one of them was a bit swollen, and another one has a slightly melted terminal. So to be on the safe side, we retired these two batteries, and we switched to a 6-battery 12V-660Ah bank, which should be better suited for our two 140W panels.

We did a factory reset on the Kid, and reprogrammed these values:

Type battery = AGM
Absorb voltage = 14.4 V
Float voltage = 13.6 V
Equalize voltage = OFF
Temperature compensation = -3 mV/°C/Cell
Absorption time = 2.5 hours

With these new values and 6 batteries instead of 8, there is a slight improvement. Voltage is now 12.8V around noon. (Note: This voltage is the one seen by the datalogger. I suppose there is some loss in the cable between battery box and datalogger, so the real voltage is actually closer to 13V)

But last year in July, voltage was much higher, like 14.5V.

So it looks like my batteries need some equalizing.

I've read many contradictory advices about equalizing AGM batteries, like:

QuoteNEVER equalize these batteries against the manufactures advice. AGM batteries are known in the industry as VRLA (valve regulated lead acid). These batteries have a stringent charge requirement to prevent a condition known as thermal runaway at high temperatures and to maintain full charge at low temperatures. Without the proper charger, these batteries have been known to explode. (sailnet.com forum)

The most useful bit of information I found is from East Penn Deka Solar Batteries:

QuoteEqualize charging not required on VRLA (AGM/Gel) as part of a daily charge setup. Based on PV applications, unpredictable recharge availability, periodic equalize may be required. (eastpennmanufacturing.com)

Unfortunately this is only for Deka batteries. Lifeline and Trojan batteries have a similar approach.

But my battery manufacturer is Enerwatt (made in China, distributed by Trans-Canada Energies here in Quebec) and there is nothing about equalization in the spec sheet of my Enerwatt WPHR6-220...

So do I need to equalize?
I guess I do...

What EQ voltage and duration should I use?
Something mild like 14.6 V for 6 hours? or something brutal like 15.5V for 2 hours?

We are planning also to add a 3rd panel, a bit like this (photoshop simulation):



This will help in summertime, to compensate the loss with the 2 vertical PVs (in case this is also a factor).

Any advice about EQ, or whatever, is much welcome!

Vic

Just a few general comments;

AGM batteries tend to like long Absorbs,   particularly if there are multiple strings of batteries in parallel.

One other common comment about AGMs,   is that for some manufacturers,  EQ or freshening charges are just long Abosrb stages,,  sometimes at a slightly higher Vabs.

But,   as you know  some slight amount of Catalyst in the battery is consumed every hour in Absorb,   so,  there is some tradeoff.

IMO,  there should essentially be NO voltage drop in the Data Logger's cable,   as essentially NO current should be flowing in that cable (knowing absolutely nothing about that logger).

Most of us always recommend to follow the advice of the battery manufacturer regarding all charge parameters.

Some/many AGM batteries have specific Limits recommended for the range of Temperature Compensation of charge voltages  ...   does the manufacturer of your batteries have such data available?

Others here know much more about AGMs than I.

Good Luck!   Vic
Off Grid - Sys 1: 2ea SW+ 5548, Surrette 4KS25 1280 AH, 5.25 KW PV, Classic 150,WB, Beta Barcelona, Beta KID
Sys 2: SW+ 5548s, 4KS25s, 5.88 KW PV, 2 ea. Classic 150, WB, HB CC-needs remote Monitoring/Control, site=remote.
 MN Bkrs/Bxs/Combiners. Thanks MN for Great Products/Svc/Support&This Forum!!

ClassicCrazy

I think I would set the equalize for absorb voltage but for a longer time than absorb - like four hours . Wonder if your system could ever get to that voltage and hold it there for that long ? 

When you said your batteries at 6.1 and 6.2 v looked normal - that is pretty low voltage if they have been on the charger .
If you really want to check the batteries next time you are out there I would bring something to be able to put a good sized load on them and see if the voltage stay up or takes a dive .

I think your new panel put at the 45 degree angle is going to help you with  your summer charging.

How many amps is the load on your system ?

Larry
system 1
Classic 150 , 5s3p  Kyocera 135watt , 12s Soneil 2v 540amp lead crystal for 24v pack , Outback 3524 inverter
system 2
 5s 135w Kyocero , 3s3p 270w Kyocera  to Classic 150 ,   8s Kyocera 225w to Hawkes Bay Jakiper 48v 15kwh LiFePO4 , Outback VFX 3648 inverter
system 3
KID / Brat portable