Questions about batteries and the midnite classic 200

Well we have finally moved out to the property and the fun stuff has begun. I bought 10 305 watt panels, 16 12v 115ah batteries and 2 3000 watt inverters. Now I was ready! Hmmmm I forgot about a charge controller. Now I have 12v batteries, 12v inverters and 24v panels so I need to convert the power from the panels to 12v. Not knowing how this could be done, I called MW&S for an answer. After a few calls I ordered a 24v 5000 watt inverter and a Midnite Classic 200  MPPT charge controller. After wiring the batteries into sets of 2 in series I have 7 24v batteries. (While drilling through the wall I took out a battery so 7pairs instead of 8. Ooops)  Now here's the thing,  I had 1840ah of storage at 12v but now with the 24 volt system and the ooops I only have 805. Also, I was told that the classic could only handle 2 parallel sets of 3 panals in series so instead of 3050 Watts of panels I have 1830 watts in use. After reading the manual for the controller, I believe it would have worked on a 12v battery bank and. I am not sure but possibly handled 3 parallel sets of panels instead of 2 which would have me producing 2745 watts instead of 1830 watts pushing our average daily output to 14kw. Anyone use this controller? Am I right? Close? Or dead wrong? Panel specs are as follows:

Rated. Power.     305 watts
Open circuit voltage 45.6v
Max power voltage.  36.9v
Short circuit current. 8.32A
Max power current.  8.94A
Power allowance range   -0/+3%
Max system voltage.   1000v
Fuse rating   15A

Any thoughts?

No idea, but i want to give you a tip. I lost 4 $169 batteries cause I never checked the water in them. They boiled out on my solar set up. It was costly.

I have heard that so many times. Good thing too because until then I had no clue that you were supposed to watch that. We check ours weekly as a result. Thanks for the input though, perhaps it will save others a good deal of money.

There are many variables to consider. We'll start with some basic calculations.

The maximum output current for your charge controller is 79 Amps. It supports battery banks with voltages of up to 72V. The wattage your controller can output is a function of your max amperage and the output voltage.

If we assume your 12V battery bank is charging at 14.4V, your charge controller can output a maximum of around 79Amps X 14.4V=  1137 Watts in a 12V configuration.

When you boost your battery bank to 24V, you effectively double your charging capability.

79 Amps X 28.8V charge voltage = 2275 watts in a 24V configuration.

The above numbers are rough, because the charge profile of lead acid is not based on a fixed voltage, but rather a voltage range. Still, the information is accurate enough to serve for system analysis.

Charge controller output of 2745 Watts is significantly higher than either calculation above; it cannot be supported in either a 12V or 24V configuration. It could be supported if you rewired your battery bank to be 36V or higher, but this entails purchasing a new inverter. Even this solution will not give you an optimized system though. For peak battery longevity, you want to charge at a rate of between C/10 and C/20; higher current will generate more waste heat resulting in lost energy, water loss in the batteries, and a reduced duty cycle. You can charge lead acid at as high a rate as C/5, but most don't for the aforementioned reasons.

Allowing for your one lost cell, your battery bank's capacity is 15 batteries X 12V X 115 ah = 20,700 watt hours. A C/10 charge rate is 2070 watts and a C/20 charge rate is 1035 watts. In either case, you have more wattage coming from your panels than is optimal for your batteries. This isn't a bad thing, because panels only perform at peak for a small portion of the day; your extra capacity buys you a steeper profile on your charge curve, whereby you reach your maximum charging current earlier in the day and hold it for more of the day. It also affords you a much more healthy charge profile in winter months.

To squeeze every watt of energy from your panels, and allowing for a C/10 charging rate, you would need 30,500 watt hours of battery capacity (305 watts X 10 panels X 10 to allow for the C/10 charge rate). But don't go shedding more dollars just yet. This is just information; it needs to be applied to your scenario before a recommendation can be made.

There is a lot to consider in calculating your optimal configuration: grid-tie or off-grid, winter/shoulder season power utilization rates, solar insulation, operating temperatures of your battery banks, typical regional weather (max days without sun), peak current draw in the home, upgrade plans, the angle of panel installation and/or solar tracking... and these are just a few. If you want to provide more detail I can help you with system design. What you have proposed with respect to system design cannot be supported by your present hardware, but there may yet be a better configuration you wish to implement. It would be a shame to waste so many panel watts or to install and be limited by a system that could be optimized in advance.

Chris Wells was faster and included one important fact. The maximum output current.

first a note: The Energy storage is measured in Watt · hours (Wh) or KiloWatt · hours (kWh)
1Wh = 1Ah (Ampere hour) · System Voltage. It does not matter whether 12V · 200Ah or 24V · 100Ah are used.

Then I mixed up some data. (removed)

When using 24V, you can still use your 12V inverters by at the cost of caution to balance the two banks. This would mean to measure them all the time and use the inverter of the bank that has more energy left…

And as Chris wrote, the capacity to current ratio is another issue.

Thanks for the info guys. Chris what info do you need? I live in the desert in West Texas so we don't have many cloiudy days to contend with. We are off grid totally. We are running 3 7.5 watt LED lights. A refrigerator that is supposed to be 349kw per year. 2 19w flat screens. A small dorm refrigerator that uses 116 watts when the compressor is running. A 5 ft chest freezer that I have not put the kill-a-watt on yet. An evap cooler that runs 250watts in the daytime and 156watts at night. And a 20 in box fan that uses 65watts 24/7.  I plan to add 3 1600watt wind turbines and buy 8 more 305 watt panels so I have 2 more 6 panel arrays. Batteries will be added as needed. I don't want to have power issues when the house is built so I am trying to work everything out now. We are trying to be conservative but something's you just got to have and once the house is built in ow our power needs will grow too. I plan to start construction of the house (Adobe) in the spring so I have time right now to save for and buy power generation equipment.

I advise that you use the midnite classic string calculator. When you enter your module specs, you can try several configurations.

I like the method of setting the max charge current on a controller to the charge rate required by your battery and letting the peak get clipped a little (rare unless drastically oversized) and using the "unusable watts" to beef up charge rate during less than peak sun. This was stated eloquently by another poster...

There is a max pv input wattage defined by midnite, obey it and all will be well.

One more thing, you can use a relay controlled by the classic to divert power or wire some of your modules into another system, even if all modules are mounted in one place. Additionally, there are pv direct appliances, ac units water pumps, rock tumblers, water heaters, almost endless ways to put all your modules to work without building another stand alone system or doing a remodel of your old one.

Always good to use up the battery you have or sell it and rebuild it entirely new than to introduce new batteries to old, unless you replace the damaged one before the others are cycled much, if new when one was damaged.

Your usage is much higher than the panels and battery bank will allow. You will need to scale your system up to support the consumption rates you mention. While there are many variables that will come into play, this is a fact we can use to start planning.

The Midnite Classic 200 supports panel voltage of up to 200V + battery voltage, and output to battery banks of up to 72V; both input and output current are limited to 79 Amps. This controller is sufficient to accept connection with your present panels and more, provided you increase the voltage of your battery bank. If you run a 60V battery bank (multiple strings of 5 batteries X 12V connected in series), it would charge at about 72V. Your charge controller could output as much as 72V X 79 Amps = 5688 watts. Allowing for the fact your solar panels can output as much as 20% more than they are rated for, you could accept input from 15 solar panels. It is possible to upsize the battery bank to 72 volts nominal, but I tend not to run equipment at it's highest rated output as the device lifespan can be shortened.

The Classic 200 supports input voltage of 200 OCV DC + battery voltage. This introduces a minor limitation with respect to panel configuration; if you use a 60V nominal battery bank, your OCV can rise to but not beyond 260V DC. The OCV of your panels is 45.6V per panel, so you can utilize five panels in series per run and remain below the peak permitted voltage. This means each string of solar will provide ~305 watts X 5 panels = 1525 watts per string. Amperage is not additive in series, so each string still peaks at only 9.94 Amps; you can connect three strings for peak output, while remaining well below the controller maximum input amperage. You might increase to four strings of solar panels; system efficiency would drop but total output would rise. You would need to determine if the charge controller can accommodate that excess input as there will be times it cannot be directed to the battery; the charge controller would either have to dissipate the heat, or increase resistance to the panels so as to eliminate the excess power. If the Midnite 200 supports this considerable overage, I would run four strings of five panels each. System efficiency will be lower, but you'll extend your battery life and be better able to deal with occasional inclement weather.

Assuming configuration with three strings of five panels, and a 60V nominal voltage battery bank, you would be getting a peak of about 4575 watts output from your charge controller. Your hardware is operating well within spec... aside from the battery bank which would need to grow.

There are losses in the controller, but also gains from panel overproduction so we'll consider losses moot. If we assume a peak output of 4575 watts and a C/10 charge rate, you need 45,750 watt hours or more in your battery bank. Each battery in your bank provides 12.6V X 115Ah = 1449 watt hours. You'll need nearly 32 batteries, but since your configuration is 60 volts, you need strings of five batteries connected in series. 35 batteries will provide you seven strings of 5 batteries each = a 60V battery bank. Your total battery capacity is 50kwh. Remember however, that a shallow draw down will extend battery lifespans, and that drawdown greater than 50% dramatically reduces cell life. This means that only 50% of your actual battery bank is usable capacity. Factoring in the losses you will experience due to temperature, DC-AC conversion, and cabling, it would be practical to consider this capacity to be about 20kwh.

Your freezer and refrigeration can be expected to use about up to 3 kwh per day. Your evaporative cooler is adding another 3kwh for 12 day use hours, and an additional 1.9kwh for 12 night use hours. Your box fan is using 1.6kwh. The LED lighting is negligible. Let's allow a conservative 2kwh for other in-house use (we assume you do not use an electric water heater, stove, dishwasher, clothes dryer, clothes washer, or any other high wattage electric appliances.) Your total consumption calculates to 11.5kwh. Your battery bank is sufficiently sized to accommodate one day of heavy cloud cover and a second day of intermittent cloud. Winter power generation would be less, but your evaporative cooler wouldn't likely be running, so consumption would offset this some.

I'd say this is about as small as you could go. Were you to add a fourth string of solar panels (20 panels in total), your output would peak at about 72V X 79Amps = 5688 watts; your ideal battery bank capacity would be 56,880 or more kwh, so you'd need 8 strings of 5 batteries each = 40 batteries. Your practical capacity for draw down would increase to 25kwh.

This may seem like a big system, but the typical home uses ~900kwh per month (at least where I live). That's a daily consumption rate of 30kwh.

There are more factors which can be considered; you have to decide if that expense is acceptable, if there are ways to reduce your power consumption, if your power consumption is likely to increase beyond that which we've accommodated, and how much supplemental electricity you can generate with the wind farm.

Wouldn't a 60v battery bank mean buying yet another inverter? Mine is 24v

I regretfully confirm that your inverter is not optimal. You can force it to work, but that too will cost you. You'll have to purchase two more charge controllers; you'll also face higher costs when you install the wind farm. Your schematic will change as the battery bank will be restricted to 24V.

The key to minimizing equipment costs rests in an understanding of amperage. Low voltage DC is high amperage DC. High amperage requires thicker wiring. Thicker wiring requires more costly charge controllers, or more of them. It means a thicker trunk line feeding your inverter. It means a larger cutoff switch. It means larger breakers for all DC equipment. It means more losses over distance. The lower the voltage of your DC battery bank, the more you're going to spend on wiring and components.

If you run a 24V battery bank, your Midnite 200 charge controller is limited to 79 Amps X 28.8V = 2275 watts output. If you run a 60V battery bank, the same Midnite 200 charge controller rises in output capacity to 79 Amps X 72V = 5688 watts. Your capacity for power generation increases to 250% of what it is at 24V. This is the Achilles heal of low voltage DC. The lower the voltage, the greater the power losses and implementation costs.

The same can be seen in wiring. If you wish to transport 5000 watts of power a distance of 8 feet (as you might with your present inverter) at 28.8V, you would need 4/0 gauge copper cable. Each wire in that cable would be a full 107 square mm in cross section area. 5% of your power would be lost as it traveled those 8 feet. Alternatively, you could transport that same amount of energy at 72V on a 4 gauge copper line with a cross section area of only 21 square mm. The weight of your copper line is reduced to less than 20% of what it would otherwise be. Thinner wiring is cheaper wiring.

Is your inverter a sunk cost, or can you recover most of it? There are ways to keep it, but you'll end up spending money in other areas to do so.

I guess I should have covered wire in the details. I am running 4ga from the panels to the controller (40 ft),  2ga from the controller to the batteries (8 ft),  00 for battery cables and 00 from batteries to the inverter (9 ft). So I think I am good on the wiring. Also the freezer is unplugged over night.

My point on the cabling may not have been clear. You need thicker wiring in all respects where your voltage is lower. This is the reason you get so much less out of a 79Amp charge controller at 24V than you do at 60V. The wiring gauge and other circuitry in the device is only sufficient for 79 Amps, and when your voltage from the controller to the battery bank is lower, the maximum amount of usable power supplied by the controller is lower as well. When measuring power availability, you have to think in watts and watt hours, not Amps. Devoid of Volts, Amps mean nothing. Watts are Amps X Volts, and thus a more complete metric for analysis.

The amount of electrical energy that can flow through a system is based on Amps times your chosen voltage. The higher your voltage, the more useful electrical energy you have to work with.

Most wiring you have is fine for your present system, but you'll be purchasing more if you scale your system to the usage level you've indicated. You need to factor that in. 2 gauge wire from the controller to the battery is sufficient to move 79 amps @ 28.8v a distance of 8 feet. The 00 wiring from the batteries to the inverter is insufficient. It's suitable for up to 3200 watts of current draw at 5% loss; your inverter is rated for 5000 watts. If you try to run the inverter near capacity for a spell, you're going to hit trouble as the cabling will overheat rapidly. If you do stay with 24V, use two lines of 00 wire instead of one, or use 0000 wire; that reduces your current loss to about 3% at a draw of 5000 watts. It resolves the meltdown risk which is otherwise pronounced (at full load). You have to respect the fact that full load could be reached intentionally or accidentally; safeguards must be in place.

Unplugging the freezer overnight will have a negligible impact on your total power consumption. It may drop consumption by a hundred or so watt hours per day; that won't change the size of the system you need.

You have a choice before you. To scale the system to your needs, you'll either have to purchase more charge controllers to overcome the obstacles of your 24V battery bank, or replace the inverter and get more out of the hardware you do purchase. In either case, you'll also need more panels and batteries.

Another factor is now coming into play. For your usage, you'll need 36-40 batteries of the size you indicated. In a 24 volt system, that's 18-20 strings of batteries connected in parallel. Balancing issues will arise. You'll need to ensure all cables from the charge controller to the batteries are of matching lengths; I would connect them to a bus bar and then connect the bus bar to your charge controller. Your run length will be determined by the least direct route the 18-20 strings of batteries take.

Your needs really push the limits of what 24V can effectively do. If you don't want to increase your battery bank voltage, perhaps you will consider ways to reduce your consumption. Anything you can run direct DC will make a notable difference; there are high efficiency refrigerators, freezers, fans, and other devices for your chosen voltage. Converting your evaporative cooler to direct DC would make the most significant difference. Note that your direct DC appliances will require heavy cabling as the amperage will rise substantially at 24V DC.

Sorry it took so long getting back to you. I am back out driving OTR to pay for this stuff lol. I am trying to wrap my head around the information provided here but a lot of it is new to me so please bare with me. My big concern here is batteries. I was thinking that I am losing storage by wiring the batteries in series. 12 @ 115ah x2 = 230ah but put them in series and I have 24v @ 115ah. Wouldn't that mean I have cut my power in half?  Now jumping to 48 volts wouldn't I need four times the batteries to store the same amount of power? I am hoping I have missed something here. And if I have, then I need to looking of a good deal on batteries. I have spent a good deal of money already in changing to 24v. I don't wanna waste more in going to 48v if it means needing 4 times the batteries for the same amount of power storage. While I do finally understand your point about wiring, wire is far cheaper than batteries.

It is a lot of information to take in. To resolve your concern, know that the number of batteries you need will be the same regardless of whether you use 24V, 48V, or 60V as your battery bank voltage.

Power is measured in watts, (volts times amps); your batteries provide the same amount of power regardless of how they are connected. What changes is the number of charge controllers (fewer where higher voltage battery banks are used), the gauge of wire (thinner where higher voltage battery banks are used), the losses from transmission of the power (losses are reduced in higher voltage battery banks), and the battery balancing issues (also reduced in a high voltage DC battery array). There are a lot of benefits to going high voltage on your battery bank, and few drawbacks. The one primary consideration is that you own a 24V inverter.

OK got it. So my next quest will be looking for a 48v inverter and more batteries.  Switching to 48v I am guessing the wire sizes I currently use will be good?  (I will likely be switching to buzz bars across the batteries themselves). According to Midnight's string calculator the best setup to still use one controller ( for now ) with all 10 panels on hand is to run 5 strings of 2 because of the hyperVOC settings. This will allow for full use of the 3050 watts rated output of the panels. With five hours of peak sunlight I will be in the 15--17 kW range which should hand the current load while I get together a new Battery bank and more panels and another controller. We are running on the generator from noon til 5 pm and the freezer and larger refrigerator are plugged into it directly and exclusively until the changes are made. The dorm fridge can handle things for now and we know the freezer can maintain for 2 days without any power so we are producing just over double our daily usage. Any ideas on a reasonably priced 48v inverter?

That's a system you will be proud to own. I have no doubt it will serve you well.

Converting to 48V will enable to use your existing wiring, and you'll get the added bonus of better efficiency.

I am very fond of OutBack inverters. I know of two 48 volt models you might consider - the VFX3648 and the GS4048A. The first provides up to 3600 watts of power and the second provides 4000 watts. Both are pure sinewave inverters; this is important as 1) some devices are incompatible with step wave inverters, 2) devices such as fridges and freezers use considerably less power on pure sine power when compared to step wave, and 3) devices tend to fail less frequently on pure sine wave as it's easier on electronic components. Inverter efficiency is a bit lower on pure sine inverters than step, this is the cost of having clean AC output. I consider the benefits of pure sine to far outweigh the efficiency reduction, but be aware of the difference as you'll see it if you compare specs between pure sine and step wave inverters.

The VFX model is stackable, so you can easily add more units if your power demands increase. It is less costly than the GS model. The unit has a couple really cool features: 1) it can start and stop your generator when your batteries get low, so you can literally run on solar as often and long as possible, while remaining safe in the knowledge that you won't run shy of power even if you aren't home. This reduces the risk of food loss in your fridge or freezer, and it also reduces wear on the batteries as the feature can be used to optimize your cycling depth (batteries last much longer if you do this). This feature will enable you to shift all of your loads to solar now, saving money and reducing generator wear even before your system expands. 2) Three units can be stacked to provide 3-phase power; you may not need this, but I do and it's awesome to see suppliers still considering this in their design. I think you can stack two units to get 240V, but you should verify that if you need 240. There may be a lower cost option, as the unit is available in a European model that provides 230V @ 50Hz. It may be possible to adapt the output of the European model or base model to provide 240V AC @ 60Hz from a single unit. If you need 240V, I'd call Outback with your needs and they'll put a tech on the phone to help you.

The GS model is more costly than the VFX, but it is grid-interactive, which means the single unit fully supports grid-tie and stand-alone usage. It's a good model if you want to sell power back to the utility. I'm not certain it matches you, but review its features if grid-tie is in your plan. One major differentiating feature does come to mind; it can handle massive dynamic changes in current needs with no measurable voltage dip. This means your well pump may fire up, but you'll not know inside as the power won't fluctuate. It's a feature of value for homesteaders who get hit with random heavy loads, you'll again want to decide if it is important to you. VFX and other manufacturer's units can handle considerable reasonable load changes without brownouts, just not the extreme shifts in power demand that apply to the GS.

There are other units out there; these are just the two I know best. If you find a different model of interest and you want to know more about it, ask away. I'm always up for a bit of research.

The FXR series has grid/hybrid control and operation, like the GS without 240v capability.

In many cases we use the GVFX inverters as off grid inverters because if a power line becomes available or new owners come into play, then it is grid interactive.

http://www.outbackpower.com/outback-products/make-the-power/fxr-series-inverter-chargers/item/fxr-renewable-series-120v-a-models?category_id=530

http://www.outbackpower.com/outback-products/make-the-power/grid-interactive-inverter-chargers/item/vented-gvfx3648?category_id=517

The GS series is comparable in price per watt, if you require 240v and or have or plan on having a 240v generator.

Either way Outback equipment is the best of the best, in my experience and is making our client's off grid systems a reliable success, without exception.

No conversation about the best equipment for RE is complete without this history lesson!

http://www.midnitesolar.com/pages/frontPage/nwHistory/history.php

I love it Chris! Joseph, im sure you know by now, that the info Chris is relaying is something uncommon, and he does it with care.

Thanks guys. That's what Permies is all about - helping each other.

I'm thinking Joseph has a solid plan now. I'm sure questions will still arise though, and I look forward to them.

Thanks for the tips on the other inverters; I'll look into them to increase my knowledge base.

Be well everyone.

Thanks guys, and you are absolutely right Chris, I will still have questions lol. And here is my next one just to make sure I understand the battery thing. I have 14 115ah batteries wired to be 7 24v sets so I have just over 19kwh and because of the 50% rule I have 9.5kwh that I can safely use. If I put my panels into 5 strings of 2, my charge controller can handle them all so I have full output of about 17kw a day. But instead of running around 110v to the charge controller I would be running around 72v. Am I getting this right? And will my 2ga wire from panels to the controller be big enough?

You have it right.

I have 14 115ah batteries wired to be 7 24v sets so I have just over 19kwh and because of the 50% rule I have 9.5kwh that I can safely use. 100% correct. A couple things are worth considering here though: 1) you extend your battery life if you don't consistently draw them down to the 50% level and 2) some power is lost as heat as it moves through the electrical lines, and as it is converted to 110V for use. This means your batteries can supply up to 9.5kwh to you, but the amount you can actually consume in appliances and fixtures will be lower. A good rule of thumb is to expect 80% usable power; you'll likely get just a bit more. So the 9.5kwh from your batteries will likely translate to about 8kwh in actual power at the appliance level.

If I put my panels into 5 strings of 2, my charge controller can handle them all so I have full output of about 17kw a day. Yes. The actual output depends on the angle of the panels, the amount of cloud cover, and the angle of incidence with respect to the sun. You are correct, but you can still optimize a bit to get more from your panels. In a rack mount configuration, it is good to enable a vertical tilt that you can manually set. As the seasons change, you simply ratchet your panels up or down. You'll gain a respectable amount of extra output for that fifteen minutes of effort every couple months.

But instead of running around 110v to the charge controller I would be running around 72v. Mostly, yes. A 36V solar panel doesn't put out 36 volts. It puts out a range, from OCV (open circuit voltage) at the high end, to max power voltage (the voltage your controller will more or less optimize to), and then less as the amount of sunlight diminishes. When calculating your string voltage and wiring requirements, it is recommended that you use your OCV, as this is the highest voltage your panels should supply. You then use your max Amperage in your calculations. This leaves you a cushion, as you'll never output both peak voltage and peak amperage. The cushion is sufficient to address periods where the sun is exceptionally intense; your panels can output more than they are rated for in these times, but they'll still be within OCV X peak Amps total output.

What this translates into is that each of your strings could output voltages as high as 45.6V X 2 (91.2Volts) or amperage as high as 8.94 Amps.

And will my 2ga wire from panels to the controller be big enough? In the configuration you outline, absolutely. Your 2 gauge wire is thick enough for you to combine all five strings near the panels, and to then use a single run of 2 gauge cable to feed to the controller. You'll end up with about 72V @ 45 Amps running through the 2 gauge cable that leads to your charge controller. It's a solid wiring configuration with plenty of excess capacity. You could also wire each string to the charge controller; this is common in high amperage configurations; you won't cause issues doing it, but there is no need for the configuration you stated. It would only be an advantage if you intend to increase the voltage of your panel strings later on. As an example, you might plan to go from 5 strings of 2 panels to 5 strings of 5 panels, in which case you could optimize your wiring in anticipation of that change.

If you have an Android phone, you can make wire gauge calculations yourself quite easily. There is an app called "Wire Gauge Calculator" which will do the heavy lifting. It's free, easy to use, and accurate. I use it when I plan electrical systems.

Can you mix battery sizes? Ex. 8 12v batteries made into 4 24v strings and 12 6v batteries made into 3 24v strings. Then all 7 strings run parallel. Will this work? What are the drawbacks?

Mixing battery sizes will result in balancing issues in the battery banks.

To run strings in parallel, you need each string to be similar in both resistance and capacity. This results in balance as the load is drawn from the battery bank, and similar balance when charge is applied. It is highly unlikely your strings will equally share the load if the batteries are as different as you suggest. As a result, some batteries would be drawn down faster than others. This leads to premature battery failure as some batteries would be drawn below their 50% threshold even while the total capacity of your battery bank was above 50%... perhaps much above. A similar issue arises with respect to charging; if the impedance in the strings differs, the controller will be unable to optimize the charge rate for all strings. Some batteries will overcharge and off-gas, even while others are still charging; if the controller continues the batteries that are off-gassing are damaged; if the controller stops, batteries with too little charge sulfate and lose capacity.

Those are just a couple examples of why you need matching batteries.  Professionals generally recommend that all batteries be purchased at the same time, as batteries that have been in use will have changed in characteristics. My experience suggests you can push the limits on this advice, provided the used batteries have not been extensively cycled. I suggest same manufacturer and part number so you can be assured of similar charge and discharge characteristics.

I would recommend having the battery supplier use an analyzer on your batteries to assess the probability of issues arising. Professional lead/acid analyzers will measure SOH (state of health) and SOC (stage of charge). If the SOH is similar on your batteries to that of the new batteries (within a few percent), you can match with relative certainty. The greater the difference in SOH, the more likely that battery will be a sore spot.

Battery analyzers all use proprietary methods to calculate SOH. You'll get useful numbers only if the same make/model of analyzer is used to test all batteries.

Chris,
I have been looking at 6v batteries and I am confused by AH rating on some.  One in particular is a Duracell brand that says 230AH at 20 hrs. I thought 230AH would be 230 hrs @ 1amp. this 20 hours thing has me lost. Any thoughts you would care to share? TIA  -Joe

Happy to clarify that for you.

The capacity of a battery is affected by the rate at which you draw it down. When you extract energy very rapidly, the battery is less efficient. What the figure you present means is that the battery supplies 230 Amp hours if you discharge it at 1/20th of its rated capacity per hour. The same battery might provide you 260 or more Amp hours if you draw it down at one Amp per hour, or only 190 Amp hours if you discharge it a rate of 20 Amps per hour. Because the battery capacity depends on the rate at which you draw current from it, there needs to be a standard. Most manufacturers rate their batteries in accordance with an assumed 20 hour discharge period.

To give you a real-life example, use TN-350 Nickel Iron batteries in my system. The 350 refers to Amp hours at a 20 hour draw down rate. I use energy at a much lower pace, so I effectively get 450 Amp hours from my batteries. By sizing my battery bank for a slow discharge rate, I can draw nearly 30% more energy from the batteries than they are rated for. Pretty cool eh?

An estimate at 1/20th draw down rate is conservative for properly sized solar configurations. Your discharge rate should be less than that, so your battery should outperform its specifications. Be aware though, just because they rate the battery for total capacity, does not mean you can use that full capacity. The rating is academic only. You can't practically plan on dropping below 50% of battery capacity without rapid degradation. As such, you should expect a 230 Amp hour battery to supply approximately 115 Amp hours of energy to your inverter.

Sebastian Köln wrote:

When using 24V, you can still use your 12V inverters by at the cost of caution to balance the two banks. This would mean to measure them all the time and use the inverter of the bank that has more energy left…

You can put one 12V inverter across half of 24V but be careful about putting a 2nd 12V inverter across the other half. One inverter would be grounded at 0V and the other would be grounded at 12V. Depending on where the grounds go and where other components in the system are grounded, you could destroy one inverter and half, and then all your batteries (in theory the inverters and charge controllers would shut down before this happened depending how good their fail safes are).

This can probably be done safely but you really need to think about the grounding on this one.

Thanks Chris,
I hate to be a pain but........

Found these for $600 but the math doesn't work for me.  That part I don't get is in red below. am I missing something? in my application I would need 6 to get to a 24v string. You know my specs, would this give me enough storage?

BTW I rewired the array so I am using all 10 panels. the only config the Classic would accept to use them all to their full output was 5 strings of 2 panels.

4KS 21PS
SERIES 5000 DEEP CYCLE
1104AH - 20HR
1468AH - 72HR
1557AH - 100H
Weight: 267LB
Measuring 4V @ 10A at this time (5-6-16)

   Dimensions (LxWxH): 15.75 × 9.75 × 24.75 inches
   Weight: 267 lbs
   Deep Cycle
   Flooded Lead-Acid
   Rope handles
   Reliable and dependable
   Dual container construction for a stronger shell
   Ten (10) year warranty for Rolls 4KS21P (4-KS-21P). First 36 months are full replacement. Next 84 months are prorated.
   This is a serious battery for large systems offering performance over a long service life.
  For a single string 1,557 AH Bank @100 Hr. rate, you will need 3 of the 4KS21P (4-KS21P) for a 12 Volt system, 6 of the 4KS21P (4-KS-21P) for a 24 Volt system and 12 of the 4KS21P (4-KS-21P) for a 48 volt system.
   Each additional string will multiply the AH.
   Real world expectations: The Rolls Surrette 4KS21P (4-KS-21P) will support a 25 amp load for 56 hours!

Your consultant appears to have a good understanding of the equipment, but could relay it differently. I would suggest Watt hours as a metric instead. Watt hours are the measure of total power in the battery bank. It's necessary to use this metric when looking at battery banks of different voltages. That's why his/her discussion in your highlighted red section doesn't seem to make sense.

Translating what your consultant said:

For a single string 1,557 AH Bank @100 Hr. rate, you will need 3 of the 4KS21P (4-KS21P) for a 12 Volt system, 6 of the 4KS21P (4-KS-21P) for a 24 Volt system and 12 of the 4KS21P (4-KS-21P) for a 48 volt system.

into watt hours, you get the following:

For a 12V string, you'll need 3 of the 4KS21P (4V X 3 batteries = 12V) and you'll get a capacity of 19,618 Watt hours based on the 100hr rate. For a 24V string, you'll need six of these batteries (4V X 6 batteries = 24V) and you'll get a capacity of 39,236 Watt hours at the 100hr rate. For a 48V string, you'll need twelve of these batteries and you'll get a capacity of 78,472 Watt hours (100hr rate).

Do you see the difference when looking at Watt hours? You can now see total capacity of the battery bank. Your devices use watts... that's a measure of power consumption. It's the measure your consultant should be using. Based on this information, we can effectively size the battery bank. But I wish to look at this differently because the calculations cannot be accurately made using an assumed 100 hour rate. Draw down rates significantly influence total battery capacity, and the above does not allow for this.

We previously calculated your daily consumption to be about 11,500 watt hours (post from 6/27/2016 12:32:27 PM). This translates into an average consumption (11,500 watt hours/24 hours) of 480 watts . That's what we use to determine the battery bank required.

We can now see what the different battery banks can supply.

The 12V pack (3 batteries in series) can't be measured at its 100 hour draw down capacity, because 19,618 watt hours [rated 1557 AH (100 hr rate) X 12.6 volts] /100 hrs is only 196.18 watts. You'll average 480 watts.
The 12V pack (3 batteries in series) can't be measured at its 72 hour draw down capacity, because 18,497 watt hours [rated 1468 AH (72 hr rate) X 12.6 volts] /72 hrs is only 256.90 watts. It's still far less than the required 480 watts.
The 12V pack (3 batteries in series) can be measured at its 20 hour draw down capacity, because 13,910 watt hours [rated 1104 AH (20 hr rate) X 12.6 volts] / 20 hrs is 695 watts. This is sufficient to supply your average draw of 480 watts.

The 24V pack (6 batteries in series) can't be measured at its 100 hour draw down capacity, because 39,236 watt hours [rated 1557 AH (100 hr rate) X 25.2 volts] /100 hrs is only 392.36 watts. You'll average 480 watts.
The 24V pack (6 batteries in series) can be measured at its 72 hour draw down capacity, because 36,994 watt hours [rated 1468 AH (72 hr rate) X 25.2 volts] /72 hrs is 513.80 watts. It can supply the required 480 watts.

The 48V pack (12 batteries in series) can be measured at its 100 hour draw down capacity, because 78,472 watt hours [rated 1557 AH (100 hr rate) X 50.4 volts] /100 hrs is 784.72 watts. The easily meets the required 480 watts.

We now know what capacities to plan on for each battery bank size (in bold above). We now halve total capacity to get usable capacity, because only 50% of the battery can be drawn down before excessive wear becomes a concern.

We can rely on the 12V (3 cell) battery bank to supply 13,910 watt hours/2 = 6,955 watt hours of usable energy to your inverter. You use an average of 480 watts, so you can expect this to power your house for 6,955 watt hours / 480 watts = 14.5 hours. Allowing for losses in the inverter and power transmission lines of 15%, your real world outcome would be about 14.5 hours X 85% = 12.3 hours. The 12V option is too small. It isn't even enough power to run your house for one day.

We can rely on the 24V (6 cell) battery bank to supply 36,994 watt hours/2 = 18,497 watt hours of usable energy to your inverter. You use an average of 480 watts, so you can expect this to power your house for 18,497 watt hours / 480 watts = 38.5 hours. Allowing for losses in the inverter and power transmission lines of 15%, your real world outcome would be about 38.5 hours X 85% = 32.7 hours. The 24V option is exceedingly trim. It will run your house for about a day and a third. This would mean trouble every time storms blow through, and during cloudy periods when you get little solar to recharge your batteries.

We can rely on the 48V (12 cell) battery bank to supply 78,472 watt hours/2 = 39,236 watt hours of usable energy to your inverter. You use an average of 480 watts, so you can expect this to power your house for 39,236 watt hours / 480 watts = 81.75 hours. Allowing for losses in the inverter and power transmission lines of 15%, your real world outcome would be about 81.75 hours X 85% = 69.5 hours. The 48V option is enough to run your house for nearly three days.

Now the actual decision on what's acceptable falls to you. The batteries should be able to supply power for the longest period you feel solar energy will be limited. For example, you might feel two days of cloud cover are the most you'll see for inclement weather; if so, the 48V 12 cell option is sufficient.

You do have the option to charge your system from other energy sources, so you don't have to consider the absolute worst case scenario when it comes to cloud cover. There's always the possibility of charging your batteries from a generator for example; a lot of people size their systems for typical weather and then use a generator for top-offs when weather is uncommonly poor. You wouldn't want to undersize and use this strategy, but for occasional unexpected poor weather, it's a good backstop.

I must also point out that there is nearly no difference between one string of 12 cells (48V) or 2 strings of six cells (24V) when it comes to the power they'll supply you. If you still intend to use the 24V inverter, simply run two strings for your battery bank.

I should also add... it's my pleasure to help you. Do not feel guilty about asking.

There are two other things you may wish to know:

1) A 4V lead-acid battery is comprised of two 2V cells. These cells have a typical voltage of 2.10V... slightly higher than 2V. That means a 4V cell is actually about 4.2V. This is why the calculations above are based on a 12V battery bank being 12.6V. Likewise for the 24 and 48V configurations.

2) When calculating how long you need your batteries to last, you start from sundown. For ease of explanation, we'll assume a 12hr day and a 12hr night.

If your batteries can carry the house for 12 hours, they can last the night and no longer. They cannot handle ANY inclement weather.
If your batteries can carry the house for 24 hours, they can last the night and one day of inclement weather, but they'll be dead by nightfall and thus unable to supply power for the second night. This means they can't handle even one day of inclement weather.
If your batteries can carry the house for 36 hours, they can last the first night, the one day of inclement weather, and the second night. The usable capacity of the batteries will be exhausted after two nights with one day between.

You can see based on how this calculates, that you must time from nightfall instead of daybreak. You have to think in 12 hrs + 24 hours. 36 hours accommodates one day of inclement weather, 60 hours accommodate two days, and 84 hours accommodate three days. Please consider this when you decide how many batteries you require.

As always I really appreciate your insight and trust me when I say it is not falling on def ears. When I had to return to work (bills never seem to stop lol) I still had the original 14 batteries and six panels up. Because of your advice here I had them shut down the power from 11pm until 7am every night and from 11am to 4pm the generator was running the house allowing the batteries to be fully charged during our peak sun hours. This was working well to keep things going until funds were available to upgrade to a new battery bank. I figured by cutting the use of the batteries to 12 hours most of which we had good sun, and running the deep freeze only while the generator was running, I could make things work for a bit while i saved for batteries. Unfortunately, my sister (my right hand when I am on the road) when on vacation with my mom for a few weeks. During this time, work was still being done on the property and things needed to be moved around a bit. One of these things was the generator. It was decided (not by me and to my horror) that for ease in switching from one power source to the other the generator would be placed next to the battery bank currently under a lean-to because I had to leave before sheeting the sides completing the shed. Still not really a big deal, but guess which way they pointed the muffler! Yup 10 inches from the batteries and aimed right at them. For 3 weeks for 5 hours every day, the generator was cooking the bank. Nobody noticed this oversight until I got home and saw it. We now have $1700 in batteries that can barely make it to 10pm after 18-20kw are producing and we should only be using about 8kw from the bank. They now have power when the sun shines and as long as they can buy fuel for the generator. I will have a 60kw battery bank in place and secured before I switch it over. We WILL NOT be mixing the new batteries with the old and I refuse to hook up the new ones until I have enough to complete the new bank. Of course I am now the biggest @ss in the project now because I threw a fit over the wasted money and refuse to install new batteries right away. We now produce almost double what we use each day and I figure based on that a 60kw bank should give us an easy 2 day reserve. The panels should support it for now and later I will add another 10 to them. Now it is just a matter of getting the most bang for the buck. I know that 2 200ah batteries while rated the same in capacity, can be as different as night and day when it comes to quality. this is why I am no longer looking a $88 batteries. I didnt go cheap with a charge controller because 10x the price was 100x the quality, and now I am starting to look at higher end batteries as well so rest assured I will be bugging you some more lol.

It's a shame about your batteries, but it does appear you have a good solution. A 60kw battery bank is right in line with a 2 day reserve. Supplemented by occasional generator use, this seems a realistic configuration for your locale. You get two hairy thumbs up from this side of the fence.

May I suggest you configure the generator to automatically start and charge when battery voltage reaches a minimum threshold? This will reduce tending, and it will also extend the lifespan of your batteries. Not every user will want to understand the technical aspects of the electrical system; many will just want it to work. This will be more able to meet such expectations. It'll also help reduce the "oops" factor that could otherwise apply if users don't pay attention to battery charge levels.

So, I have decided on switching to a 48v system. I want to use bus bars instead of cables as jumpers between the batteries. I found some copper bars but dont know if they will work. Here are the stats:

110 Copper Rectangular Bar, Unpolished (Mill) Finish, H02 Temper, ASTM B187, 1/8" Thickness, 3/4" Width, 12" Length (99.9% Pure Copper)

Any advice?