combining unlike panel array's.

dellsworth
dellsworth Registered Users Posts: 20
Hi,
I went to a new customer site yesterday. The house is a bit underpowered and their renter is unhappy. The owner want's me to give a proposal for an upgrade. My problem is that I've yet to do an upgrade combining unlike panels.

The current system is 10 80 watt Kyocera panels going into a home made combiner box that apparently can't be expanded. I want to double the size of the array but am not really sure how to combine the two arrays. Obviously I want to use newer 24V panels - probably 4 225W panels. It's currently a 24V system with a morningstar pro charge controller. I forgot to look it up and see what amperage load it can support. I have a used Xantrex c60 I could probably replace it with if I have to or get a new MPT I suppose although the customer is very tight with money.

Just as a side issue I'll post here a picture of the homemade combiner box. It's got rectifiers instead of blocking diodes. I didn't pull it apart so am a bit confused how it works. All the 10 red leads at the top are 24 volts. The panel conduit has one conduit for each pair but 4 wires in each conduit. You can see the 5 conduits coming in at the top of the box. I'd have to pull the facing board off to see how they're connected at the back but I didn't bother figuring I'd have to bypass it to combine with the new array downstream (but before the charge controller) anyway. The #10 red wire branching off at the bottom goes to a lightening arrester. Obviously the #8 red and #8 white at the bottom go to the charge controller. It's a pretty piece of work. I wish I knew who did it but the owner died and his wife knows nothing.

Thoughts and suggestions are welcome.


Attachment not found.Attachment not found.

Dan

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: combining unlike panel array's.

    The Imp of the array is probably around 22 Amps--And the largest MorningStar Pro PWM controller is 30 amps--So you will need to add/replace the controller with a larger one.

    I would suggest you use a DC Current Clamp Meter and measure the current flow through each panel.

    Also, I am not sure why 10 diodes... If these are 17.4 volt Vmp panels, then two in series for each string--then would only need 5 diodes.

    Check the voltage drop across each diode. It is possible for them to fail shorted, and this would not be "legal" typically for NEC/UL/NRTL. Normally, two diodes in series would be required in event one diode fails shorted, a second is still there to block reverse current.

    It sort of looks like each block of 5 panels is "diode blocked" then connected together in parallel, then each block is connected together in series. While it will work (solar panels are OK to connect in parallel because they are, more or less, current sources and will share current based on the amount of sun on each panel (unlike batteries which are very sensitive to voltage and resistance miss-matches).

    What is the Battery AH capacity. What is the daily load (Watt*Hours/Amp*Hours)? What inverter and is it on 24x7, search mode, etc.

    Review 24 volt panels. Make sure they are really around Vmp~35-38 volts or so--Many "24 volt" panels are really Vmp~29-30 volts or so and not really recommended to parallel connect with a 35-38 Vmp array.

    When combining panels in series, Imp should match to within ~10%. Same for parallel connections, Vmp should match within ~10%. For a PWM charge controller--Vmp really only has to be above the Vmp-array-minimum for your battery bank (35-38 volts suggested). Higher Vmp panels can be used as long as the controller is not over voltaged... Too high, and you waste a lot of power with PWM controllers (lost power from Vmp-array>>Vbatt-charging).

    Review current wire gauge for voltage drop. Try to make sure no more than 3% drop for array. If the array is a fair distance fromt the battery bank/charge controller--perhaps look at replacing with MPPT controller. Of course--if you your Panel Vmp miss-matches are too great--You may be forced to have two controllers (original PWM, and a second MPPT). Additional costs of two controllers would probably be a no sale.

    Review Charge Controller to battery bank connection. Short/heavy wire... I would suggest no more than 0.2 volt drop (for 24 volt battery bank) for accurate battery voltage feed back to the controller (and maximum current from charge controller). If parallel battery strings--I would suggest using a DC current Clamp and DMM to measure current sharing under heavy load/charging and battery voltage across each cell/battery (look for too high/low readings/differences). Also check specific gravity of all cells.

    It looks like the Ground Wire for the solar panel/mounting frames may go through the combiner box. If so, I would highly suggest that the frame ground wiring be separated and go straight to earth in a short/few curves/no corners method to local ground rod or building ground rod if close to foot of array. Rewire ground wire to outside of building (run down exterior wall) to ground. If you have chance of direct strike from lightning, it is really a bad idea to bring the lightning ground wire into the home (especially into the "center" of the home. Lighting wants to flow outwards (i.e., two lightning ground cables should be at opposite sides/corners of the home).

    Also review battery grounding (or not) and AC grounding (or not--especially if typical MSW inverter).

    Review loads, review battery bank against loads, inverter capacity vs loads, review array size against loads/amount of sun, and battery bank 5%-13% rate of charge.

    Review battery history--It sounds like there is a good chance the battery is damaged by now. And who knows the battery capacity vs daily loads vs solar array sizing (balanced system design).

    Review any backup generator/AC battery charger against battery AH capacity and generator rated wattage (not too large genset, waste fuel) and required loads (charging only, shop loads, etc.).

    So--My list:
    1. If they don't have a battery monitor (Victron another good brand), I would suggest installing one now--Allow them to measure their loads and see how the current array is performing for their needs. Also--I think the percent state of charge display would be much easier for tenants to read/monitor (an off grid rental?).
    2. Check current through each panel and make sure each one is working.
    3. Rewire with traditional fused/breakered combiner box with two panels in series per fuse/breaker (safety, remove diode drop--could be 1.0 volt per diode or ~2.0 volt from array--causing under voltage array on hot days/during Equalization). Then parallel.
    4. Rewire safety/frame grounds for lightning considerations.

    Anyway, those are the major points I would look at.

    Very interesting--this is an Off Grid Home with tenants? Does the owner live on site and use/share same system?

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Re: combining unlike panel array's.

    from what i can see that is 5 paralleled kc80 pvs in series with 5 paralleled kc80 pvs. rather than put fuses at each pv output they put a blocking diode. the blocking diodes should not be necessary with most charge controllers today, but the fusing i would do. that would make the battery bank 24v and new pvs paralleled to this will need to match the voltages from the kc80 arrangement. the 60 cell pvs will not match the voltages needed by the present arrangement. most 72 cell 24v pvs should match the voltage, but due to current limitations on the present controller this will need its own controller and then it shall go to the battery bank.

    the battery bank may also need revamped. if the present batteries are old then replace them all or be able to switch battery banks.
  • TheBackRoads
    TheBackRoads Solar Expert Posts: 274 ✭✭
    Re: combining unlike panel array's.

    If it were me, I'd add the desired array on a new MPPT charge controller and leave the existing setup. Maybe upgrade that CC as well depending on age...depending on overall system health.. Just my thoughts..
  • dellsworth
    dellsworth Registered Users Posts: 20
    Re: combining unlike panel array's.

    Hi everyone,

    Sorry I'm late with this. I thought I'd get an email for any posted messages. I finally looked and found all your great responses. I think I fixed my profile.

    So I've read everything in detail and then did some more research on the forum as thebackroads suggested. It looks to me like the best option is to just create a duplicate system and put the positive leads from the charge controllers to the same battery post. I think I have to NOT use an MPPT. I think it'd get in a twist with the Morningstar and definitely get in a twist with another MPPT. It certainly seems easier than trying to find panels that won't cause problems. The only added cost will be another CC (cheaper than an MPPT) and a bit more wiring - and the combiner box which all amounts to not much compared to the headache of coming up with something closely compatible down here in the remote south of Mexico.

    There was lots of great information from all of you. I really enjoyed the speculations from Bill and Niel about the combiner box. If I do this job I'm going to pull the back and see how it's all connected. I generally follow NEC guidelines but here in Mexico I don't have to follow much of anything and there's no oversight. Even the grid-tie systems I've done CFE doesn't even look and they have no guidelines. It's just paperwork. This, however, IS off the grid. The original owner was zapotec/Mexican and his wife American. She hated Mexico and when he died I don't think she waited a week to get out. She just tries to find people to rent the house until she can sell it. Most people don't last long. They don't get an explanation on how to live there with a weak solar system and try to live there as if it's a normal house. A couple of women have been afraid of the remoteness and left all the lights burning all the time. If she could get away with selling it without telling anyone that the solar system really isn't robust enough she'd do it. However it's a small community and I think the word is out so she's going to have to bite the bullet to get a sale. When they lived there about half the summer (the rainy season) they'd run a generator to supplement. In 2010 they replaced the battery bank with 20 115 amp/hour batteries. 1150 amp/hour. When I looked at it the red light (discharged on the morningstar) was flashing. The kid that was watching the place (no renters at the moment) said it has been flashing most of the summer since they lose power now and then and it takes days to get the bank charged again. I called him last night and he said the bank was only at 23.6 at 8pm and he'd turned off the inverter before going home in the afternoon for lunch! I've not done a full power-off check but I'd say they've got some fried batteries from too many 20% cycles and not getting fully recharged. I've had pretty good luck extending banks by yanking the bad ones and rebuilding the banks and might do that here if she lets (pays) me. It would probably work better with less batteries anyway. I don't know how they decided to put in that big a bank with such a small array.

    I did look at the voltage drop but it was a bit confusing and need to look again. I think as Bill suggested there was an added drop because of the diodes. I took readings on both sides and was confused by the differences. At one point I was reading 26 at the conduit and 25.1 at the controller and then below the diode I was reading 25.5. I didn't think about the diode resistance and need to confirm. They're running #8 (as you can probably tell from the picture) but my clamp-on only showed 19 amps at 25 (take your pick) volts bright sun overhead and it's only about 4 feet combiner box to charge controller.

    Oh. I think they're actually rectifiers - not diodes. I looked up rectifiers and how they work DC to DC and apparently they work like diodes and maybe with less resistance? Maybe one of you guys who know such things could clarify.

    The inverter is a Xantrex trace 24V but didn't notice if it was a 1500 or 2500 but it looked like the DR series (or the newer version of the same) so modified.

    GREAT suggestion about grounding. The panels are probably okay if they're generating 19+ amps but I can check them. When I got there they were generating 17+ and I rotated the array to face the sun - no clouds - slight haze. They use a generator but on a switch. They shut off power from the inverter when running the generator and vice-versa. Whoever set it up didn't understand about the Trace.

    Thanks to you all. I really enjoy having you gurus available to help out basic get by guys like me. Send me a note if anyone wants a place to hang out on a Mexican vacation. I've lots of room - and sun :-)

    Dan
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: combining unlike panel array's.

    Dan,

    The Battery bank at 1,150 @ 24 volt is just too large for that array... The nominal 5% to 13% range we recommend for a "happy battery bank" would be:
    • 1,150 AH * 29 volts charging * 1/0.77 panel+controller derating * 0.05 rate of charge = 2,166 Watt array minimum
    • 1,150 AH * 29 volts charging * 1/0.77 panel+controller derating * 0.10 rate of charge = 4,331 Watt array nominal
    • 1,150 AH * 29 volts charging * 1/0.77 panel+controller derating * 0.13 rate of charge = 5,631 Watt array "cost effective maximum"

    Assuming a 5/10% rate of charge with an 800 watt array (current) or plus 900 watts more would be:
    • 800 watt array * 1/29 volts charging * 0.77 panel+charger derating * 1/0.10 rate of charge = 212 AH @ 24 volts
    • 800 watt array * 1/29 volts charging * 0.77 panel+charger derating * 1/0.05 rate of charge = 225 AH @ 24 volts
    • 1,700 watt array * 1/29 volts charging * 0.77 panel+charger derating * 1/0.10 rate of charge = 451 AH @ 24 volts
    • 1,700 watt array * 1/29 volts charging * 0.77 panel+charger derating * 1/0.05 rate of charge = 903 AH @ 24 volts

    So, those would be the numbers I would aim at.

    It appears your array running at >17.6 amps at noon would seem to indicate the equipment is OK (at least when charging near dead batteries).

    Hot weather and batteries near 29 volts when in absorb may suffer because of voltage/wiring and diode drops. A nice MPPT controller with Vmp >> 35 volts would be nice for hot weather. But--You probably have problems with the owner putting that much money into the system.

    Those "rectifiers" are regular silicon diodes with ~0.9 volt drop at 6 amps (as far as I know--hopefully this links to the data sheet).

    Assuming you get lots of sun and not too much monsoon/"weather"--You may get close to 5 hours of sun per day... An 800 watt array would generate (AC inverter) around:
    • 800 watts * 0.52 system derating * 5 hours of sun = 2,080 watt*hours per day.

    So--reducing the battery bank (by finding the good batteries) is a good thing.

    There may be a point here where the owner decides what is a "sufficient" off grid system for the home. At 800 watts, I would be really tempted to drop back to a 12 volt battery bank, a MorningStar 300 watt TSW Inverter, and a 45 to 60 amp PWM controller (I would suggest MPPT--But, again, the cost may make this a hard sell).

    Or, up to 24 volt battery bank and a Rogue 12/24 volt 30 amp MPPT charge controller. And find a "good"/smaller 24 volt Inverter.

    Are they trying to run an AC refrigerator off the system? That will set the lower end of the inverter at 1,200 to 1,500 watt (typical).

    A battery monitor would be great -- But again, more money, hard sell.

    Getting a kill-a-watt type meter so you can demonstrate AC loads... Review lighting--Perhaps some 24 VDC LED lighting would help (leave on, timer, motion detector (12 volt version), etc.) so people feel safer (I am more of a motion detector type guy--leaving lights on over night is more of an attractive nuisance vs deterrent in my humble opinion).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • dellsworth
    dellsworth Registered Users Posts: 20
    Re: combining unlike panel array's.

    Bill,

    I knew the bank was too big but didn't know the exact formula. I've sized them traditionally according to load requirements at the same time as configuring the panels. I appreciate the formula. On the .77 figure for derating... Is that the same for a mppt as for the PWM? The figure assumes no more than 2% loss in wiring?

    Your .52 for system derating is right on for here. I've been using it for several years and it's worked perfectly.

    I also love light sensors. They make perfect sense for remote solar homes. I have them all over the place. The newer sensors work well with compact florescent. I wish I had some but even the old ones don't burn out bulbs too frequently - which is what seems to happen.

    If I put in separate stand-alone systems and tie them at the battery bank wouldn't I run into problems with two mppt's? I had another client that we'd upgraded to grid-tie with an outback GT upgrade - which requires at least 100 amp/hours of batteries to function. The existing system was using a xantrex c40. Outback designs their GT units to "talk" with their MX. It was a nightmare of literally weeks of screwing around with the configuration to make it work with the xantrex. I can't say enough bad things about outback in this particular instance. They gave me a support guy who I think had never even seen an inverter and refused to pass me to a real engineer. I finally gave up on them and played with it until it worked. Anyway both the xantrex and the GT wanted to manage the batteries and for various reasons in the GT it has to be involved - you can't simply let it ignore battery charging (even though it's an option). It simply doesn't work with that configuration using grid-tie. I suppose this would be easier without grid-tie but after that experience I'm gun shy of two controllers managing the same bank. It'd be nice to have something like a master and slave arrangement like outback does for the inverters. But no one will do that I suppose for cheap charge controllers.

    All in all it seems any way other than simply putting the two "stand-alone" systems to the same battery bank is fraught with matching problems - whether they be controllers or panels.

    They have a "real" home - not really designed for off the grid. I suppose it was driven by the American wife. So yes they've a fridge and a bigger one - maybe 16CF or even 18. They've a couple of 1/2HP water pumps - one that drives a pressure tank for water. Obviously that one isn't on much but they have a cistern pump also for getting water run off. I run mine during the afternoon after the batteries are charged and the kid (caretaker) said they do also. They've no led's but all compact florescent. They've a microwave. One renter that was really concerned called me so I went over to explain how to "manage" his consumption by showing him what various appliances were pulling with my clamp-on. He had been using a big (old and inefficient) desktop computer as a server and had a printer and laptop also. All on at the same time - all the time. He stopped doing that. No one will buy or live there without a reasonably robust system. It'll have to be a 1500-2400 watt inverter.

    Hey people want stuff. I did a system for a woman that insisted she wanted a commercial 220 electric stove/oven. It more than doubled the price of the system just for that appliance alone :-) She didn't care.

    Thanks so much for your support - particularly with the formula. I'm eventually going to be educated.

    Dan
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: combining unlike panel array's.
    dellsworth wrote: »
    Bill,

    I knew the bank was too big but didn't know the exact formula. I've sized them traditionally according to load requirements at the same time as configuring the panels. I appreciate the formula. On the .77 figure for derating... Is that the same for a mppt as for the PWM? The figure assumes no more than 2% loss in wiring?
    Dan,

    The number is the "default" from PV Watts... Roughly 81% for solar panel derating from STC and wiring/other losses. The other is 95% eff for a MPPT GT inverter.

    For PWM controllers, they only "care" about the Imp-array only (no down converting as in the MPPT controller--PWM is current in - current out).
    • 14.5 volt battery / 17.5 Vmp-array = 0.82 = 82% losses because of "no down converting" with PWM

    So, for a "hot array" in summer derating from STC to PTC of ~82% with MPPT controllers (p=V*I where Vmp is temperature sensitive and falls with increasing panel temperatures) vs the, more or less constant current (Imp actually rises a little bit with higher panel temperatures--but not really enough for us to "care" about), ends up with virtually the same output in summer.

    In winter (near and sub freezing), you can get 10% to 15% more power because of the lower panel temperatures (and higher Vmp)--But I usually do not take that into account because +/-10% is about as close as any generic estimates are going to be. Between hardware/panel variations and long/short term weather effects--Being closer is just not worth taking into account (and winter production is so much less because of shorter hours of sun--10%-15% more of shorter/less power days, is not a very large number for most people).

    If people get a little more power during the 3 months of fall/winter/short days, it just reduces generator runtime a bit. And for most people, the random multiple days of poor sun (weather) is really driving the generator runtime anyway.
    Your .52 for system derating is right on for here. I've been using it for several years and it's worked perfectly.

    Glad to hear it... Others off-gridders here use 1/2 or 50% efficiency too--virtually the same as 52%. Probably a bit on the conservative side (assumes 80% eff for flooded cell batteries and 85% for inverters plus 77% for panel+controller deratings)--But close enough for most people that they are happy with the results.
    I also love light sensors. They make perfect sense for remote solar homes. I have them all over the place. The newer sensors work well with compact florescent. I wish I had some but even the old ones don't burn out bulbs too frequently - which is what seems to happen.

    LED lamps are probably the best now--Low power, on/off cycling does not affect life, and "instant on" during cold weather.
    If I put in separate stand-alone systems and tie them at the battery bank wouldn't I run into problems with two mppt's? I had another client that we'd upgraded to grid-tie with an outback GT upgrade - which requires at least 100 amp/hours of batteries to function. The existing system was using a xantrex c40. Outback designs their GT units to "talk" with their MX. It was a nightmare of literally weeks of screwing around with the configuration to make it work with the xantrex. I can't say enough bad things about outback in this particular instance. They gave me a support guy who I think had never even seen an inverter and refused to pass me to a real engineer. I finally gave up on them and played with it until it worked. Anyway both the xantrex and the GT wanted to manage the batteries and for various reasons in the GT it has to be involved - you can't simply let it ignore battery charging (even though it's an option). It simply doesn't work with that configuration using grid-tie. I suppose this would be easier without grid-tie but after that experience I'm gun shy of two controllers managing the same bank. It'd be nice to have something like a master and slave arrangement like outback does for the inverters. But no one will do that I suppose for cheap charge controllers.

    Generallly have not read about very many "parallel" charge controller issues with charging a generic off grid system.

    My recommendation is to "home run" the wire (short/heavy cable) from controller to battery bank. And not to "daisy chain" the wiring from battery to controller "A" then to controller "B", etc.... From what I understand, the variation of voltage caused by the different controllers due to voltage drop (PWM mode, MPPT "sweeps", etc.) can "confuse connected controllers (enter absorb/float too early?). Home running back to the battery bus should reduce controller to controller interference. (note: one poster here found that "desuflators" created enough electrical noise on his system that it reduced the output of a pair of Outback FX/FM MPPT charge controller's daily output very significantly--by 10-20% or so).
    All in all it seems any way other than simply putting the two "stand-alone" systems to the same battery bank is fraught with matching problems - whether they be controllers or panels.

    If you are trying to use a GT inverter "back feeding" a OG inverter--Then I could see issues depending on how you control the GT feedback. SMA should do this well (native design with Sunny Island product). A few Off Grid Inverter vendors now shift the 60 Hz frequency when the battery bank is full (Xantrex XW, some Magnum?, etc.). The shifting of the frequency "knocks" the GT inverter off line until the frequeny is back at 60Hz.
    They have a "real" home - not really designed for off the grid. I suppose it was driven by the American wife. So yes they've a fridge and a bigger one - maybe 16CF or even 18. They've a couple of 1/2HP water pumps - one that drives a pressure tank for water. Obviously that one isn't on much but they have a cistern pump also for getting water run off. I run mine during the afternoon after the batteries are charged and the kid (caretaker) said they do also. They've no led's but all compact florescent. They've a microwave. One renter that was really concerned called me so I went over to explain how to "manage" his consumption by showing him what various appliances were pulling with my clamp-on. He had been using a big (old and inefficient) desktop computer as a server and had a printer and laptop also. All on at the same time - all the time. He stopped doing that. No one will buy or live there without a reasonably robust system. It'll have to be a 1500-2400 watt inverter.

    Sounds like a recipe for failure. Unless somebody is willing to install a pretty large solar array to support the large battery bank. It takes a special renter to understand the limitations of the system and money available for hardware/panels/batteries.

    And if the owner were to install/upgrade the system--I am not sure how much money they would get back from the "investment". It would be better for the buyer to do the design/investment directly--Then they have skin in the game (power vs $$ spent vs conservation). But--if the property will not move without "reliable"on-site power, some sort of money will have to be spent.
    Hey people want stuff. I did a system for a woman that insisted she wanted a commercial 220 electric stove/oven. It more than doubled the price of the system just for that appliance alone :-) She didn't care.

    Yep--power is a highly personal choice. And we here attempt to educate so that people can make the decisions they need for their home/lifestyle. I have a next door neighbor that has an electric stove because they are scared of natural gas (electricity is not cheap here in California).
    Thanks so much for your support - particularly with the formula. I'm eventually going to be educated.

    Dan

    Again, you are very welcome. And let us know how everthing works out. And feel free to add your experience to help others who are looking for answers for their questions. It is nice to have folks with lots of "real" experience out there provide real world experience on the forum.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: combining unlike panel array's.
    In 2010 they replaced the battery bank with 20 115 amp/hour batteries. 1150 amp/hour.
    That's another problem with the system. I think you are describing 10 battery strings in parallel. That is bound to fail soon even if you have adequate charging power.

    One string of large celled batteries is optimal. Two strings is acceptable. All else is trouble. When charging parallel strings of batteries, it is difficult to make the current flow into each string evenly.

    Also, one string of batteries has half as many cells to measure (voltage and spec. gravity) and maintain (water, check connections).

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • dellsworth
    dellsworth Registered Users Posts: 20
    Re: combining unlike panel array's.

    vtMaps,

    I agree 10 strings in parallel is not good. I done a number of off grid installations but none with banks quite that big. I've several with 6 and 8 strings (my own is 8) with customers I've had for 6 to 9 years. Bigger batteries are harder to come by here and expensive. Of course in this particular case we all realize the bank is overbuilt. However my experience with these 115 amp/hour "maintenance free" batteries is that they last 3 1/2 to 4 years and, after rebuilding the bank, they usually add a year. If one doesn't mind rebuilding the bank as often the cost benefit comes out way ahead. The cost of these 12V 115 amp/hour batteries down here is about $105 US dollars INCLUDING delivery on site. Compare that to high end batteries and the cost per amp/hour over time works out for the cheap ones with more strings. It's cheaper to replace them 3 times compared to once for a 15 year battery.

    I admit I've not checked higher end battery prices here in the last couple of years but up to now that rational has worked well. It only takes a couple of hours to swap out a bank of multiple batteries (or rebuild it) and the customer can get into a system for a LOT cheaper. You do have to monitor your bank a little more and yes I've noticed in all the banks with more strings the batteries charge and discharge inconsistently and you get varying voltage levels over time (even after equalizing) so you might have to swap them around every couple of years. I personally would love to have some 15 year batteries but the economics has never worked out. It doesn't hurt to occasionally go out to a customer and fine tune the bank. They feel you're keeping an eye out for them and you get repeat business, recommendations, and upgrades. At the price of these batteries overbuild the bank by a couple of batteries. Then at 3 to 4 years or so throw out the worst ones and in the process you'll swap them around. I've always gotten at least another year. That's also been my logic in using 24V systems. It gets painful to throw out 4 but 12V systems are difficult to be efficient. It's a compromise. However I disavow any configuration or design for the system we've been discussing in this thread.

    Dan

    LTH owned by Johnson Controls but manufactured in Monterrey, Mexico. They bought out LTH Mexico a few years back and the batteries are surprisingly robust - for the price.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: combining unlike panel array's.

    Dan,

    Do you think it would make sense to have the customer buy an inexpensive DC current meter so they can check charging/discharging current per string (make sure all strings are sharing properly)?

    At least in the US, Sears sells a "Good enough" DC current clamp/DMM for ~$60 each.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • dellsworth
    dellsworth Registered Users Posts: 20
    Re: combining unlike panel array's.

    Hi Bill,

    I've done that for multimeters rather than getting fancier monitors back when charge controllers didn't have anything but leds. In fact in some installations in indigenous communities in the Lacandona I was giving them away to the local authorities when I fixed systems or installed new batteries for people to borrow as well as gave a 2 hour class in solar management. It hadn't occurred to me with clamp-ons since mine were kind of expensive. I'd thought you needed RMS ones to get good readings for DC systems. However it makes sense for simply doing comparisons between two strings. Any errors in a cheap clamp-on will be the same for both. Thanks for the good suggestion.

    Dan