Help me finalize my off grid system please

Collapse
X
 
  • Time
  • Show
Clear All
new posts

  • solar pete
    replied
    Hi hammick, thanks for the news on your system. Big WOW on the beautiful location, reminds me of when I was driving around rocky mountain national park last year, awesome country, cheers

    Leave a comment:


  • Logan005
    replied
    Fantastic, Looks great! Love the building. How far off grid is this place? Looks like a nice location.

    Leave a comment:


  • hammick
    replied
    OP here with an update. My system is fully installed and so far I am very happy. For the seven days (1-26 through 2-3) experience I have using my system everything seems to work great. My living quarters isn't complete yet but I was running my travel trailer off the system (led lights, water pump, refrigerator, heater fan, etc) and I was down to about 80% SOC each morning. It was pretty cold and the heater was cycling a lot. I'm guessing when the living quarters is complete I will see about 85 - 90% SOC each morning (wood burning stove only for heat).

    One of the fully sunny days I actually saw the CC putting out more watts than the panel wattage. I'm assuming this is because it was so cold. The day I left I was using my 230v well pump a lot to winterize the trailer. In the early afternoon full sun the CC was putting out 30 more watts than the well pump was drawing. Batteries were staying in float. That's good news for the potential of adding a washing machine for use during sunny days.

    The batteries were in float when I left and I set the bulk, absorb and float voltages to 52.8v which is the low end of the float spec for the Trojan L16-REA batteries. I shut down the inverter. I still need to fuse the positive and negative cables at the battery terminals.

    Had a couple cloudy days and my 2400w inverter charged the batteries just fine. It's only 120v so the charging is 50% vs. using a 240v genny but it worked well enough I sold my monster 7000w contractor genny to my builder.

    Thanks for all the great advice.


    Last edited by hammick; 02-22-2016, 08:00 PM.

    Leave a comment:


  • Mike90250
    replied
    stratification due to low charge current.
    With L16's in their tall cases, the electrolyte will tend to stratify out to different layers. When there is enough AMPS in the PV array, to charge the batteries hard enough to really gas the batteries, the bubbles mix the layers up, and there is no problem. But with a over size battery (under paneled) you don't get enough charge current to stir the layers up, then battery life suffers.

    Leave a comment:


  • hammick
    replied
    Originally posted by sensij
    If you are primarily using the 1710 W array as described earlier to do the charging, I think water loss will be less of a risk than stratification due to low charge current. Again, I'm calculating C/13 at most for all but a few hours of the year, but clearly modeling it differently than some might.
    Thanks. If I am having problems getting the batteries charged when we are staying up there I will have three more panels installed.

    Leave a comment:


  • sensij
    replied
    Originally posted by hammick
    OP here. I will be using the Trojan L16RE-A batteries. 325ah. People have reported the RE batteries use less water for whatever reason. I will also install water miser caps and set the absorption back when I leave. Hopefully the water use will be minimal.
    If you are primarily using the 1710 W array as described earlier to do the charging, I think water loss will be less of a risk than stratification due to low charge current. Again, I'm calculating C/13 at most for all but a few hours of the year, but clearly modeling it differently than some might.

    Leave a comment:


  • Mike90250
    replied
    Originally posted by thastinger
    .... his batteries are going to see in excess of 35A at times, which I figured would boil off his water levels too quickly between visits.......
    Not going to happen. The batteries are going to bed daily, full. When the sun rises, there is at least a 2 hour ramp-up time as the sun angle improves. By the time the sun is optimum, those full batteries will, again, be full, and not accept the potential 35A charge. Maybe if there are a couple cloudy days, and the sky clears exactly at noon, then you might get 35A , for 10 minutes, as the batteries fill, the amps will fall off rapidly.

    When the cabin is occupied, and batteries low at night, then the morning-noon charge may approach the limit, but that's a different situation.

    For the idle months, absorb times can be set way back, since the batteries are not in "cycle mode", but more like Float duty.

    Leave a comment:


  • lkruper
    replied
    Originally posted by J.P.M.
    A very wise engineer once told me that rules of thumb are good for designing thumbs, but since thumbs have already been designed and seem fit for purpose, such rules are less than optimum for professional engineers, or most anyone else, as a serious choice in design. Maybe as a 1st cut elimination aid, but not much else.
    I have been told that engineers carefully calculate everything to a few extra decimal places and then when they are finished, double it for good measure

    Leave a comment:


  • J.P.M.
    replied
    Originally posted by sensij
    What you observe for your system is totally consistent with PVWatts, using the same parameters I described earlier... it shows a south facing 45 deg tilted array in VT (edit: or Portsmouth, VA, as it turns out) as reaching STC power in winter. I don't know the specific orientation of your system, but as tilt increases, extra power beyond what the base model shows is entirely believable when there is snow cover and the albedo increases. NREL's SAM gives the ability to tweak albedo to improve the model accuracy in those situations. An 18 deg tilt isn't going to see much of that, but since it sounds like the OP is designing for mostly summer / fall use, the flatter tilt is not necessarily a bad design. His system will not produce more than 30 A, and almost always, much less.

    Discussions in the grid-tie section of the forum frequently move beyond simple rules of thumb and dig deeper into the consequences of specific array locations and orientations. I think the off-grid discussions could sometimes benefit from that... rules of thumb get you close, but when it comes time to spend thousands or tens of thousands on equipment that will fail prematurely if not used correctly, a little extra work is worth the effort. That is not meant as an attack on you... everyone who contributes to helping others is doing some good, although sometimes more well-intentioned good than actual good. It is mostly a warning to those who come seeking advice... rules of thumb are frequently not good enough to design an optimal system, and if the advice you get is based solely on those rules and is not specific to the circumstances, it may not be right. If the only advice being received is based on those rules, it may be because not enough information is being provided.
    A very wise engineer once told me that rules of thumb are good for designing thumbs, but since thumbs have already been designed and seem fit for purpose, such rules are less than optimum for professional engineers, or most anyone else, as a serious choice in design. Maybe as a 1st cut elimination aid, but not much else.

    Leave a comment:


  • sensij
    replied
    Originally posted by thastinger
    My array is pretty optimally placed for my Lat and on a 35 degree cloudless day my panels exceed their rated capacity by as much as 8 percent. I've never has an off-grid system where it gets as cold as where the OP is located nor tilted at 18 degrees but my feeling is that he could certainly approach his rated output.
    What you observe for your system is totally consistent with PVWatts, using the same parameters I described earlier... it shows a south facing 45 deg tilted array in VT (edit: or Portsmouth, VA, as it turns out) as reaching STC power in winter. I don't know the specific orientation of your system, but as tilt increases, extra power beyond what the base model shows is entirely believable when there is snow cover and the albedo increases. NREL's SAM gives the ability to tweak albedo to improve the model accuracy in those situations. An 18 deg tilt isn't going to see much of that, but since it sounds like the OP is designing for mostly summer / fall use, the flatter tilt is not necessarily a bad design. His system will not produce more than 30 A, and almost always, much less.

    Discussions in the grid-tie section of the forum frequently move beyond simple rules of thumb and dig deeper into the consequences of specific array locations and orientations. I think the off-grid discussions could sometimes benefit from that... rules of thumb get you close, but when it comes time to spend thousands or tens of thousands on equipment that will fail prematurely if not used correctly, a little extra work is worth the effort. That is not meant as an attack on you... everyone who contributes to helping others is doing some good, although sometimes more well-intentioned good than actual good. It is mostly a warning to those who come seeking advice... rules of thumb are frequently not good enough to design an optimal system, and if the advice you get is based solely on those rules and is not specific to the circumstances, it may not be right. If the only advice being received is based on those rules, it may be because not enough information is being provided.
    Last edited by sensij; 10-23-2015, 03:06 PM. Reason: wrong location, fixed

    Leave a comment:


  • thastinger
    replied
    I understood the OP intends to leave the system unattended for a few months? Working off that assumption and knowing that his CC is going to reset at midnight, his batteries are going to see in excess of 35A at times, which I figured would boil off his water levels too quickly between visits. I did not pay attention to the 18 degree tilt so it wouldn't be as bad as I originally thought but would still be over 30A I would estimate. My array is pretty optimally placed for my Lat and on a 35 degree cloudless day my panels exceed their rated capacity by as much as 8 percent. I've never has an off-grid system where it gets as cold as where the OP is located nor tilted at 18 degrees but my feeling is that he could certainly approach his rated output.
    I agree with Sensi on the real world aspects as my system output rarely exceeds 80% capacity on hot summer days with plenty of sun so extra panels would be good during those times, but in the winter my system cranks.

    Leave a comment:


  • sensij
    replied
    We are starting to talk past each other, so I'll take one more stab at this. Let's start with what we know:

    1) OP is in S. Central Montana (I'm using Livingston weather file for modeling in this post)
    2) The 1710 W (6 * 285 W panel) array is mounted 180 deg azimuth, 18.5 deg tilt.
    3) OP has a 60 A high quality MPPT charge controller. CC clipping is not on the table in this discussion.
    4) The PV output of the grid-tied array defined above can be modeled by PVWatts, using "Premium" panels, roof mount, 8% loss. These parameters fit real-world data for healthy grid-tied systems very well, and a good MPPT CC could have close to the same efficiency.
    5) The OP was considering a 48 V, 208 Ah battery, and was told that 1710 W was too much array for this battery.
    6) We don't know what the daily energy requirement is, but for the sake of conversation, let's say we start the day at 50% SOC.

    Let's start to make some assumptions about the state of the battery. I'm sure these numbers can be nit-picked, and I'll be happy to adjust them as needed to improve the credibility of this thought process. I think they will be close enough to make the point.

    7) A battery at 50% SOC has an OCV of 48 V.
    8) OP's battery at 100% SOC has 48 V * 208 Ah = 9984 Wh of capacity (20 hr discharge rate) (call it 10000 Wh for the sake of rounding).
    9) OP's battery at 50% SOC has 10000 Wh / 2 = 5000 Wh remaining of its capacity.
    10) The round-trip efficiency is about 75%. To fully replenish 5000 Wh of capacity, 6667 Wh of energy are required.
    11) The bulk stage of charging is the only stage in which the PV is able to deliver power without constraint.
    12) The bulk stage of charging ends at about 85% SOC... that is when it shifts to adsorb and becomes voltage limited.
    13) To get from 50% SOC to 85% SOC (bulk stage charging), 3500 Wh / 0.75 = 4667 Wh of energy are required.
    14) The battery voltage at the transition from bulk to adsorb is 59 V.
    15) During the bulk stage, as the 4667 Wh are being delivered, the voltage will climb linearly with energy from 48 V to 59 V (I know it isn't really linear, but I don't think the non-linearity screws this up)

    Time for some pictures. First, scanning through the PVWatts model output, I looked for a cool, clear day with high peak power. May 24 fit the criteria, although there were a bunch of days in the April / May timeframe that would have worked for this. If the 1710 W array is truly oversized for this battery (because the C rate is too high), a day like this should show it right? If not, what type of day would?

    PVWatts Hourly.GIF

    Focusing on just the model of May 24, the grey lines below show when 4667 Wh have been generated, at 12:10 pm. Since it is in bulk, and the CC is not clipping, we'll assume that all of that energy is going towards charging the battery (with losses accounted for in the round trip efficiency).

    PVWatts - Test Day.GIF

    We know the power being delivered during bulk, and have guessed the voltage increases linearly with the energy delivered between the known 48 V and 59 V endpoints. With power and voltage defined, a plot of what the voltage and current during bulk might look like that day can be generated (P = I * V)

    PVWatts - Test Day IV.GIF

    Now that the system has gotten to adsorb at 12:10 pm, the voltage is held at 59 V and the current is allowed to taper back off as the SOC increases from 85% to 98% or so. The point has been made many times in this forum... during adsorb, the ideal case is to let the battery take whatever current it wants at that charge voltage, but with PV systems, usually the supplied current drops off faster than the natural decline in current drawn by the battery, and the system ends up undercharged. For that reason, having a larger array is better, as long as the current doesn't exceed C/8 in bulk for an extended period of time.

    You can see in this model that the maximum current observed in bulk is 25.2 A, or C/8.2. From this, I conclude that in the situation that the OP described, a 1710 W array is sized well to support a 48 V, 208 Ah battery, and is not oversized, as another forum member suggested. Supporting the 330 Ah battery the OP ended up with is going to be a stretch, and careful maintenance / supplemental charging will be required.

    Leave a comment:


  • Sunking
    replied
    Originally posted by sensij
    59 V may be too high, but 48 V is certainly too low. At what point in the charge cycle would you be able to put leads on the battery and measure 48 V?
    When they are discharged. When you apply a charge to a battery, it is not voltage, it is current. The voltage is only a product of the OCV of the battery plus IR voltage.

    Example if you have say a 48 volt AH battery the Internal Resistance is around .04 Ohms. So with a 30 amp current supply on a battery with a OCV of say 48 volts is only going to rise to 48 + (30 amps x .02 Ohms) = 48.6 volts. That means your panel power was clipped from 1710 watts down to 30 amps x 48.6 volts = 1450 watts. Had you used a 40 amp controller, you would be able to use full panel power all the way through the Bulk Stage 1710 watts / 48 volts = 35.7 amps which is the constant current phase where the fastest and largest charge is obtained. Use the Bulk Voltage as your voltage and you screw yourself out of large percentage of power. The only time you would use full panel power is the last few minutes of the Bulk stage as the battery voltage rises to 1710 watts / 30 amps = 57 volts. You just wasted a lot of time, money, and energy. You would be better off just chunking 200 watts of panels into the trash as that is essentially what you are doing. using a high voltage.

    Always use Nominal Battery Voltage or worse case for all calculations. You are sugar coating it, only fooling yourself and cheating your clients.

    Leave a comment:


  • sensij
    replied
    Originally posted by Sunking
    Sensi you do nbot use Bulk cost-off voltages. Use nominal or else you will be driving you CC into cut-off.

    1710 watts / 48 volts = 35 amps
    59 V may be too high, but 48 V is certainly too low. At what point in the charge cycle would you be able to put leads on the battery and measure 48 V?

    The voltage at the time charging begins should be something around 50 Vdc if the system is being run properly. How long will it take for that voltage to rise to the charge voltage (which I hope we agree is ~59 V)? If you are charging from the grid, a generic curve suggests 3-4 hours. With solar, probably longer, because you aren't getting full current right away. This leads to the point I am trying to make. Ignoring everything else for the moment, even if you had an array rated for 1710 W of charge power, you won't get that until you are several hours into the day, when the sun is high, and you won't get that for very long. Several hours into the day, even at the slower charge in the morning, the voltage will still have risen above 50 V... maybe not all the way to 59, but using 48 V in the calculation is totally misleading.

    I appreciate that rules of thumb can take what is a complicated combination of factors and simplify it enough to explain in a forum like this. Really, though, if the rules of thumb have convinced someone (or someones) that a 1710 W, 18 deg tilt array in Montana is too big for a 48 V 208 Ah battery, I think something is going wrong. The 330 Ah that the OP has now is not really going to be supported well by this array (if the batteries were in good condition), and the 420 Ah array he was ready to buy would have almost certainly been chronically undercharged or stratified unless they were closely monitored and maintained.

    Leave a comment:


  • Sunking
    replied
    Sensi you do nbot use Bulk cost-off voltages. Use nominal or else you will be driving you CC into cut-off.

    1710 watts / 48 volts = 35 amps

    Leave a comment:

Working...