MPPT - 55.8V array to 12V battery bank

[Walt Er]

Hi -

I'm trying to maximize solar panel voltage to minimize wire size from the panels to the controller, but run 12V batteries to power DC appliances.  I'm not sure how to calculate the amps that will get into my batteries through the MPPT controller, but I'm guessing something like this:

The panel is 200W, 55.8V.  If the MPPT controller is 97% efficient, that brings the panel power down to 194W.  If the MPPT controller matches that 194W to my 12V battery, it should be putting 16 amps into my battery, even though the panel is rated at 3.59 amps.  If I put two of those panels together in parallel, I can get a little more than the 30 amps that I need to charge my batteries at 3 hrs insolation (90 ampr hrs per day of usage).

If this is true then I only need 2 panels, compared to the four minimum I was going to need by various attempts to match the array voltage to the battery voltage.  Am I doing something wrong?

[ghurd]

A lot will depend on the battery capacity.

The 3 hour insolation makes that more important.

Just because the PVs can make it, dosen't mean the batteries can accept it.

http://www.fieldlines.com/story/2007/4/28/232143/114

A "12V" is charged at about 14.4V.  That's part of the problem.

MPPT is only going to gain about 20~25%.

G-

[Flux]

Yes I very much agree with this. The mppt will give you a gain in efficiency especially in cold weather. It may also be justified for reducing the cable cost.

Normally I would have thought that 400W was about on the low limit to justify the extra cost.

Don't get too excited, it will be rare that you see 200W from a 200W panel. As Ghurd said the battery voltage will be above 12v by a fair bit and this knocks the current down so your 16A is most likely going to be nearer 12A.

The final stage of the charging process can not be rushed, batteries just won't accept it.

Flux

[bob g]

the 97% efficiency is in my opinion very optimistic as well

you might consider ~92-93% as a more realistic number for an mppt controller.

200watts of panels rating is factored on clear skies, perfect alignment, mild temps

etc.. so i would expect maybe 150-175watts output to be more real world output

and then when the panels are fairly close to alignment with the sun and clear skies

average year round output.

i am tbinking you will be doing well to see a 10% gain using mppt with 200watts rated panels going into a 12 volt bank, overall average.

maybe a bit more if you are very careful and lucky

in my opinion on small systems, the money might be better spent on more panels instead of an expensive mppt, unless one is planning on growing the system and going to something like 48volts nominal in the future.

bob g

[Walt Er]

OK - I'll use 14.4V volts as my battery voltage when calculating amps from the panel, and use 93% efficiency for the MPPT controller (although the one I'm looking at claims 97%).  I read the previous post that ghurd linked to, though, and I have a question about matching the panel to the battery.

Ghurd, you listed this formula: solar amps equal to battery AH / 20 or 25.

My new solar amps based on the changes above are 26 amps.  Does that mean that my ideal battery bank has only 650 AH (26 X 25)?  If I use 90AH/day I'll be at 28% draw down in two days without sun, and 69% draw down in five days without sun, which is pretty common here in NH in the winter.  Originally I was planning on being able to go five days with only a 20% drawdown (2200 AH).  Am I reading that formula correctly?  

[electronbaby]

Im not sure about that formula, but it appears to me that if you want 90Ah/day with 3 hours insolation, then you should be looking at a larger PV array, MPPT or not.

[Walt Er]

Alright, let me get this straight.  

Starting with rated Watts - say 4 125W panels, 17.3V, 7.23A wired in series...

If everything was perfectly efficient and real life equaled rated life...

Using an MPPT controller I should get Vmodule/Vbattery*Imodule amps into my 12-volt battery bank.  69.2V/13.5V*7.23A = 37A.  I used 13.5V as sort of an average between 12 and 14.7.

Now, I've got to factor in the inefficiences:

Real life panel Wattage - 80% of rated

Controller efficiency - 93%

So my 37A becomes 37*.8*.93 = 27.5A.  

That's about 20% higher than I'd get if the four panels were in parallel, assuming the same 80% "real life" reduction, so that jives with some of the earlier comments.    

Are those reasonable percentages, and are there other inefficiencies I haven't considered?  

Thanks for lending your experience to this newbie - Walt

[scottsAI]

Walt Er,

Full sun with MPPT will give you around 193watts / 14volts = 13.7 amps Not as you calculated. (assuming everything works right, stated right etc.) Several times improvement over direct connection to battery, well worth doing. Else can the panel be rewired for 12v output (or cut in half) would help, still not achieving MPPT potential .

Adding a tracker can be very beneficial, several nice designs out there DIY not costly.

!!!Important discussion on Large batteries!!!!

Solar power is expensive so are LARGE batteries.

Lead Acid batteries are 91% efficient charging when state of charge (SOC) is below 80%.

SOC Above 80% efficiency drops to zero as 100% SOC is reached. Battery is fully charged.

By 90% SOC charge efficiency maybe less than 50%.

By 95% SOC charge efficiency maybe less than 25%.

If you expect to get 5 days only discharging 20%, then your working in the 90% SOC most of the time.

Several factors affect charging efficiency, SOC, battery temperature!, Battery voltage and charger.

The above numbers can be improved... another discussion. (not by a huge amount)

Lead acid batteries are made to be used, using only the upper few percent the batteries may last for many many years, you also spent a LOT of money for them. To keep the large battery charged you will need a solar panel 2-4x larger! And the battery is 4-5x larger than needed.

Over all not a good deal to over size the battery.

Charging a Lead acid battery fully is also stressful, getting above 95% requires a finishing charge raising the temperature stressing the battery!

If you NEED the battery to last forever because you believe the world is going to end (or some other like reason) they by all means over size everything with the understanding in cost today. (Storing a dry battery can last many years... (my choice)).

Reducing the battery so by the 5th day ends SOC=20%, will stress the battery, if occurs frequently like a couple dozen times a year consider using a larger battery. If not then I would not size it any bigger.

Just a thought, by starting out with the smaller battery, save the money you would have spent, you have enough to buy 4 more sets! by the time all are warren out, I bet it will be longer than the time the over sized battery set would have lasted. And you saved 2-4x on the solar panels or generator run time.

Many people with over sized batteries must run a generator frequently to keep it charged, solar panel array is not large enough to keep their loads and battery fully charged. They do not understand why this is happening. Now You know.

Hope this helped!

Have fun,

Scott.

[BigBreaker]

If the goal is to have the battery last forever get a Ni-Fe battery.  With minimal care they will outlive the owner.  In my opinion, that is more sensible than keeping dry lead acids on hand and unused.  If civilization doesn't end and we all have a Mr. Fusion running in our basements... well the nickel in those batteries will still be worth a fair fraction of their original cost.

[ghurd]

It's not some fancy theorem, just a seat of the pants, rule of thumb, that I use.

It seems to work well for me.  

Like Scott says, too much battery is no good and not enough is no good.  I am not afraid to take them down to 12.2V every day.

The 20~25% gain would be about max.  I would figure closer to 10% average.

A couple things will work in your favor.

You won't go 5 days with "No Sun".  It may not be bright, but you will get some charging everyday.

MPPT will put more amps into a lower battery than a fuller (is that a word?) battery, given the same 'sun'.

The 650AH seems reasonable to me.  And that is 3 pairs of 6V golf cart batteries from Sam's Club / Costco, though I would probably go with 4 pairs.

Might look into 2 pairs of L16 or L16-HC if the wallet can handle it.  Interstate is less expensive than Trojan.

It will make a lot more power in summer than the 3 hour insolation.

G-

[Walt Er]

Thanks for all the great battery info.  What I get from it is that I can size my battery to be at 20% SOC by the end of day 5, as long as this doesn't happen too frequently (couple dozen times per year).  

By this statement, I see that you are assuming a 96.5% controller efficiency:

"Full sun with MPPT will give you around 193watts / 14volts = 13.7 amps Not as you calculated." (Also note that I changed to four 125W panels halfway through the post).

But what about the full sun part.  The panels will only produce their peak power in full sun, perfect conditions, which is not all the time.  I thought I had already factored in the rarity of perfect conditions by using only three hours of insolation, but should I knock down the rated panel wattage by some percent also?  

Thanks - Walt

[scottsAI]

Walt Er,

Welcome, hope it was help full.

Solar for a given location is rated in full sun hours.

I live in Michigan, which is rated at 4.5 full sun hours a day. AZ is 9.

The rating is based on average over the year and weather, fixed panels.

Look around, charts for this.

The reality we have daylight for 16 hr in summer, 8-9 hr in winter.

Morning sun is weak, yet solar panels generate power during this time charging the batteries, current starts out low, reaching near peak current by 11am, stays at peak for couple hours then declines.

Tracking helps by keeping the full sun reaching the panel, can add a couple hours of near peak power a day, further north the more helpful tracking and MPPT is. (and south of equator same thing). In Michigan tracking adds about 30% power collected a day, equal to power added by MPPT (based on many assumptions, standard panels being the first one).

When I do calculations I show all of it, reduces confusion of what I'm doing, helps other to learn by seeing the example!

Have fun,

Scott.

[scottsAI]

BigBreaker,

Very good point!!

Not many suppliers of Ni-Fe batteries, may cost more than the over sized battery.

Remember to keep more electrolyte around to replace it every 10 years.

One slight down side to Ni-Fe battery is the charging efficiency is 80%, not changing over life nor state of charge (anybody know differently?). Another source suggests 70%.

Or should I say the efficiency is an advantage if you want to keep the battery topped off?

Have fun,

Scott.

[Walt Er]

Thanks everybody,

You all helped me get a much better grip on this than what I could get from the simple sizing worksheets provided by various vendors.

Based on your comments to up the wattage, get more realistic about MPPT, and properly size the battery, I think I am going to go with something like this:

Three 175W, 35.4V panels in series through an MPPT controller to put about 29amps into four L-16's (740 AH total).  

The 106V at the panels will allow me to use small (14AWG) wire for my long run to the cabin, and the 12V battery will power my daily DC load of 83 amp hrs.  I'll use a 12V/1500W inverter for my corded power tools.

Feel free to comment, heckle, or ignore at your leisure.  

Thanks again for all the help.  This site is quality.

  -Walt -- Going off-grid before snow falls....

[Walt Er]

OK before anyone beats me to it, the voltage is too high for the charge controller (at least the one I was looking at).  Voc at freezing temperatures would be too high.  Back to the drawing board. -Walt

[BigBreaker]

A bit slow to charge and discharge, not especially efficient but they certainly have a long cycle life even when fully discharged.

www.beutilityfree.com/nife.html is a source that imports them from China.

Most of the electrolyte is KOH with a dash of LiOH, so it wouldn't be too tough to store a lifetime supply as dry powder.

[ghurd]

A lot of people like the Outback MX-60.  Certainly worth a serious look.

The #14 wire has me wondering.

Any less wire losses in MPPT makes more power to the battery.

Going up to #12 would be cheap.  Even #10 is not much more money.

I use #12 for a 50W panel.

When it is finished you will understand what you have, and how it works!

G-

[scottsAI]

Walt Er,

Size wire to keep losses below 2%.

Calculator bottom of the page:

http://www.powerstream.com/Wire_Size.htm

The longer the run the thicker the wire.

Solar panels are expensive, power loss in the wire becomes expensive.

Yes, AL can be used. Make sure you have the right termination for it, etc.

My last comment about batteries!-)

Buy the cheapest you can, see how long you can make last. (proper care)

If you have not taken care of batteries before, you do not want to experiment on expensive batteries, could be costly. More than one person has ruined a good battery with poor maintenance.

Strongly consider the inexpensive automatic watering systems. Last look cost $6/cell (with shopping).

My golf cart batteries cost $57 each, have 8. three years and counting.

Many claim to get more than 7 years from cheap golf cart batteries.

Better batteries may last longer, yet need 14 years due to cost ( ratio).

I used cost/watts to compare batteries (watts stored = aHr x volts)

12v x 83ah = 1kwhr per day, not much power!

Consider bus bar vs cables.

Have fun,

Scott.

[Walt Er]

FYI,

It was the Outback I was looking at.  They have a tool that calculates Voc at various temperatures for various panels to see if the MX60/FM60/FM80 can handle it.  

The link is:  http://www.outbackpower.com/forum/viewtopic.php?f=7&t=2900

Then click on "String Sizing Tool".

The record low temp where I'll be installing the system is -25 degrees F, which made the Voc too high for these controllers.  

I have found some 180W and 195W panels with lower voltages, so I can string three together in series and still fall well within the limits of the controllers even at very cold temps.  

-Walt

[Walt Er]

Yeah, it's not much power, but we're going to try to make it work.  Using all DC, it's one efficient freezer, a few lights, a laptop, a phone, and a radio.  We'll have a generator to re-charge the batteries on days when I use a heavier load like power tools, and a seperate dedicated solar pump for the well (or a hand pump - haven't decided yet).

I was using 3% wire loss, but I'll take your advice and use 2%.  I'll think about the cheaper batteries too.  

Once the system is built and running, I'll post my set-up from my solar powered laptop and let everyone know how it's working.  -Walt

[scottsAI]

Walt Er,

Good luck hope things work out!

Well

Have you looked into Rain water catchment?

Some states could be a problem, others OK.

Looked at NM, 2100sqft roof with 5,000 gal tank, supplies water 50gal/day for 4 people.

Overall much cheaper to do than a well, unless hand dug.

Have fun,

Scott.

[tecker]

There's no reason why a bucking converter wouldn't squeeze the amps from the copper coil boosting is another story .