First solar system inquiry

Options
Markab
Markab Registered Users Posts: 18
Hi,
I want to start by first thanking the moderators and others here who
have taken the time to help so many with their solar set-ups. I
guess I'm now one of those in need of advice and direction.
I'm in the early stages of developing my first solar system. I'm
coming at it from a hobby perspective and plan to use it as an off grid
set-up to meet most of my basements electrical needs. They include the following
along with their power consumption requirements (obtained by using a watt
meter)

20" TV with cable box (90w)
Bar fridge (100w 6:1 duty cycle with 400w compressor surge)
Laptop (55w)
3 15w compact fluorescent lights (45w)

I wish to power each of these items for 4hrs per night/7 days a week
(weather permitting), and not draw more than 50% of the batteries charge.

I'm looking to buy a Rogue MPPT 3024 charge controller and probably
a 1000w inverter (suggestions welcomed). I plan on buying two
Trojan AC Series 12 vdc, 175 ah lead acid batteries. Now this brings
me to my solar panels and equipment placement. The panels will be mounted
on the roof of my shed in the backyard and I will be running a 30ft.
length of 6 awg wire back to the house where the batteries,cc,and inverter
will be housed. Now I'm sure you're all wondering whats the specs on my solar
panels are. Well, that's my first big dilemma. I will be constructing my panels
from scratch using pre-tabbed, B grade, .5v 1.8w cells. I'm looking to
construct 4 panels but can't decide if I should go with 24v or 12v panels.
24v will have less line loss (back to the house) and slightly better cc
efficiency. 12v panels will be smaller and easier to assemble. Here is my
proposed panel layouts

12v: 36 cells in series = 18v at 3.6a x 4 in parallel= 18v @ 14.4a (259.6w)

24v: 68 cells in series = 34v at 3.6a x 4 in parallel= 34v @ 14.4a (489.6w)

(I hope I got my math right...)

I'm inclined to think I'm going to need to go the 24v route to keep my
batteries properly charged. I would really appreciate any insight that
you can offer. Oh, and I live in Toronto, Ontario, Canada.

Many thanks,
Mark

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,479 admin
    Options
    Re: First solar system inquiry

    Regarding your loads--If you have a Watt*Hour meter (like a Kill-a-Watt meter)--you can measure your Watt*Hour loads and better plan your power needs.

    Guessing your average loads over 4 hours:
    • 90 watts * 4 hours = 360 WH per day == 20" TV with cable box (90w)
    • 100 watts * 1/6? * 24 hours = 400 WH == Bar fridge (100w 6:1 duty cycle with 400w compressor surge)
    • 55 watts * 4 hours = 220 WH == Laptop (55w)
    • 45 watts * 4 hours = 180 WH == 3 15w compact fluorescent lights (45w)
    • =====================
    • 1,160 Watt*Hours per day average load
    Using PV Watts program, Toronto Canada, assuming 1kW worth of solar panels (even number to estimate output), 0.52 Derating Factor (off grid inverter and flooded cell battery bank):
    "Station Identification"
    "City:","Toronto"
    "State:","ON"
    "Lat (deg N):", 43.67
    "Long (deg W):", 79.63
    "Elev (m): ", 173
    "Weather Data:","CWEC"

    "PV System Specifications"
    "DC Rating:"," 1.0 kW"
    "DC to AC Derate Factor:"," 0.520"
    "AC Rating:"," 0.5 kW"
    "Array Type: Fixed Tilt"
    "Array Tilt:"," 43.7"
    "Array Azimuth:","180.0"

    "Energy Specifications"
    "Cost of Electricity:"," 0.1 Can$/kWh"

    "Results"
    "Month", "Solar Radiation (kWh/m^2/day)", "AC Energy (kWh)", "Energy Value (Can$)"
    1, 2.69, 44, 0.04
    2, 3.51, 52, 0.04
    3, 3.98, 63, 0.05
    4, 4.68, 69, 0.06
    5, 5.29, 78, 0.07
    6, 5.48, 75, 0.06
    7, 5.51, 76, 0.07
    8, 5.23, 73, 0.06
    9, 4.96, 69, 0.06
    10, 3.61, 55, 0.05
    11, 1.95, 27, 0.02
    12, 1.97, 32, 0.03
    "Year", 4.07, 714, 0.62
    Lets assume you will run the system 9 months of the year (toss out the lowest 3 months)--We get 52 kWHrs per month per 1,000 watts of solar panels (average February). Or:
    • 52 kWH/month * 1/30 days per month = 1,733 Watt*Hour per day per 1,000 watts of panels minimum 9 months of the year
    You use 1,160 Watt*Hours per day:
    • 1,160 WH per day / 1,733 WH per day per 1,000 watts of panels = 669 Watts of panels minimum for 9 months of year
    Regarding your inverter--Your fridge may take more than 400 watts to start (most what/amp meters are not quick enough to "see" the starting surge). For a full sized frost-free refrigerator, you should be looking at a minimum of 1,200-1,500 watts of inverter. (note, I did not include the inverter stand-by losses--some are pretty high and need to be included in the original power calculations).

    Now, for the array Vmp (maximum power point voltage)--To properly charge a 12 volt battery bank, the rough Vmp minimum needs to be Vmp>2volts+Vbatt-charging... or:
    • Vmp>2 volts + 15 volts = ~17 volts minimum panel rating
    Note that Vmp falls with high ambient temperatures, and if this was in a very hot location, you may need even higher than 17 volts for properly battery charging.

    Now, there are two major types of solar charge controllers. PWM (pulse width modulation) and MPPT (maximum power point tracking).

    PWM are cheaper controllers but become less efficient the higher Vmp is with respect to Vbatt (i.e., a 34 Vmp panel connected to a 12 volt battery bank will only be 50% efficient)--so for PWM, you want your Vmp in the 17-19 volt range for efficient operation.

    MPPT are more expensive but are actual power converters. They will efficiently down convert from High Voltage / Low Current of a solar array to Low Voltage / High current to charge the battery bank.

    All About Charge Controllers
    Read this page about power tracking controllers


    Your choice with controller type you will want to use. MPPT controllers are certainly more efficient and flexible for use with larger arrays (roughly larger than 200-400 watts). PWM are much less expensive.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Markab
    Markab Registered Users Posts: 18
    Options
    Re: First solar system inquiry
    BB. wrote: »
    Regarding your loads--If you have a Watt*Hour meter (like a Kill-a-Watt meter)--you can measure your Watt*Hour loads and better plan your power needs.

    Guessing your average loads over 4 hours:
    • 90 watts * 4 hours = 360 WH per day == 20" TV with cable box (90w)
    • 100 watts * 1/6? * 24 hours = 400 WH == Bar fridge (100w 6:1 duty cycle with 400w compressor surge)
    • 55 watts * 4 hours = 220 WH == Laptop (55w)
    • 45 watts * 4 hours = 180 WH == 3 15w compact fluorescent lights (45w)
    • =====================
    • 1,160 Watt*Hours per day average load

    Hi BB,
    To clarify, the bar fridge will only be on for four hours per day (it only has beer in it). I just looked at my watt meter that has been monitoring the fridge for exactly 24h. It indicates .42 kWh. If I divide this by six (4 hours of operation) it comes to .07 kWh or 70wh. So then my wh per day should be 1,160 - 330 =
    830wh per day ? Thanks for you help.

    Mark
  • BB.
    BB. Super Moderators, Administrators Posts: 33,479 admin
    Options
    Re: First solar system inquiry

    Mark,

    Yep--that will be the number you want to aim for... Note though, you may want to cycle the bar fridge through your 4 hour cycle--if the contents (beer/etc.) are pre-chilled, then it should be close to your estimate. If the fridge needs to draw down the temperature--it could be easily running 100 watts for 4 hours.

    You can adjust your needs---I have all of the basic assumptions there. Say, this is only a "summer time" use--then obviously, you can pick another "minimum month" to build out your array for.

    By the way, make sure your DYI solar array is not over any flammable stuff (roof, ground cover, etc.)... You may be using wood/plastic in the construction and it is not unknown for these types of panels to catch fire during normal operation.

    Also, you should fuse (or breaker/combiner box when you have more than two solar arrays in parallel) each string with a series fuse roughly 2x the Isc (short circuit current) to limit the string current(s) in the case of an internal short circuit).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Markab
    Markab Registered Users Posts: 18
    Options
    Re: First solar system inquiry
    BB. wrote: »
    Using PV Watts program, Toronto Canada, assuming 1kW worth of solar panels (even number to estimate output), 0.52 Derating Factor (off grid inverter and flooded cell battery bank):

    Hi BB,
    Thanks for your help. Can you please explain the 0.52 derating factor for me. Does it mean by using an off grid inverter and flooded cell battery I'm loosing almost half of the power generated by my solar panels ?!?!
    Is there any way to improve this value? Perhaps by using certain batteries, high end inverter, etc? Thanks again for your help.

    Mark
  • BB.
    BB. Super Moderators, Administrators Posts: 33,479 admin
    Options
    Re: First solar system inquiry

    Well, PV Watts uses 0.77 derating for solar panels (roughly 80% derating, and ~95% derating for grid tied inverter), add 80% for flooded cell batteries (or 90% efficiency for AGM type), and ~85% for the inverter:
    • 0.77 * 0.80 battery * 0.85 inverter = 0.52
    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Markab
    Markab Registered Users Posts: 18
    Options
    Re: First solar system inquiry
    BB. wrote: »
    Well, PV Watts uses 0.77 derating for solar panels (roughly 80% derating, and ~95% derating for grid tied inverter), add 80% for flooded cell batteries (or 90% efficiency for AGM type), and ~85% for the inverter:
    • 0.77 * 0.80 battery * 0.85 inverter = 0.52
    -Bill

    Hi Bill,
    I'm a little confused. Is a watt not a watt? Do you mean that a PV (solar) watt is not equivelent to say, a wind generator produced watt? Where does the
    .77 come from? Thanks for helping a simpleton ;)

    Mark
  • BB.
    BB. Super Moderators, Administrators Posts: 33,479 admin
    Options
    Re: First solar system inquiry

    I should add, that this is a very rough rule of thumb and is intended to be somewhat conservative...

    There is a Derating Help link in PV watts that shows how they take certain values into account (I don't agree with all of them--but they do average out to be pretty accurate for an average season of power collection). The PV Watts program is based on average weather (included rain, clouds, smog, etc.) over some 20 years or so of measurements. Coastal/Marine areas are many times hit by humidity/fog/marine layers.

    Of course, your panels need to be mounted in full sun (or at least 9am-3pm or so)... If you are under trees, on the shady side of a mountain valley, etc.--it can badly hurt your output. Solar panel need full sun to operate efficiently. Any shading during the middle of the day can dramatically reduce solar PV panel output.

    There are lots of variables... Solar panels on MPPT charge controllers lose output power as they get hot--so mounting them on roofs with 5-6 inches of air space underneath or on rack mounts above the roof, above ground, or on poles all help to keep panels cooler. In very cold climates (below freezing), the panels can output 10-25% more power than the above estimate would show (including snow/water reflections, etc.).

    Flooded cell batteries tend to be less efficient as they age--new FL Cell batteries may start out 90% efficient and AGM's may start out in the 95-98% efficiency range. Also, how you operate the batteries (keeping them near 100% full, running lots of equalization, etc.) can also reduce efficiencies. And some batteries designed for very deep/tough cycling--like forklift batteries--will be on the less efficient side of the scale.

    Idling inverters can take from 6-20+ watts (depending on design and operating mode used).

    MSW (Modified Square Wave) type inverters can lose upwards of 20% of their output energy running inductive loads (like electric motors--fridge, etc. motors run hot) vs TSW (True Sine Wave) types.

    You can have lower losses by running large gauge wire (1-2% vs 3% maximum recommended).

    In the end, we try to give you estimates that will layout a system that will pretty much meet your operating needs. And in any case, most people tend to grow their loads over time and designing a system to be 2x larger than your current needs is not a bad thing to do either.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • BB.
    BB. Super Moderators, Administrators Posts: 33,479 admin
    Options
    Re: First solar system inquiry

    Basically, a Solar Panel Watt is a "Marketing Watt"... Clean panel, light directly overhead, Panel at ~70F. Exposed to an artificial sun only for a few moments.

    Panels are relatively cool and generate more power. Panels can be as high as 63F/35C over ambient temperature in full sun, no wind, and mounted near flush with a roof.

    Also, not everyone gets 1,000 Watts per Square Meter on a sunny day. Atmospheric Conditions can also dramatically affect the actual amount of power collected.

    Assuming around 81% "efficiency" for just the solar panel itself is a good starting point for most people. My peak daily array output is around 2,700 watts for a 3,500 watt array (and a Xantrex 3kW Grid Tied Inverter):
    • 2,700w / 3,500w = 0.77 on a "typical" clear and cool day...
    Do I see higher peaks at times--yes. Do I have lots of day with lower peaks--yep--lots of them.

    In the end, estimating a system's output to within 10% of its "true output" at your location is probably as close as one can expect. Any closer is probably pure luck.

    Also, not all charge controllers accurately log their output power... Some models seem to always estimate 5-10% higher than reality. So even measuring your output to 2% accuracy is sometimes a roll of the dice too.

    And since your collection is always based on the weather and available space for charge in your battery bank--you will either have to install a few more panels or use a bit more gasoline/diesel to keep your battery bank happy...

    Because, no matter what is predicted by PV Watts or how much load you think your system should support--monitoring your battery bank State of Charge is critical to a long and cost efficient battery life.

    Almost everyone has killed their first battery bank (me included). And almost everyone over estimates the amount of power the solar panels actually provided and minimize their true loads--Leading to "deficit charging"--Never fully charging the battery bank (above 90%) and running below 75% (or even below 50%) state of charge--which leads to sulfate hardening and reduced battery capacity over the months/years of under charging.

    Read these couple of battery FAQ's and plan your system/operations around keeping your batteries happy... Your wallet will thank you later. ;)

    Deep Cycle Battery FAQ
    www.batteryfaq.org

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • icarus
    icarus Solar Expert Posts: 5,436 ✭✭✭✭
    Options
    Re: First solar system inquiry
    Markab wrote: »
    Hi BB,
    Thanks for your help. Can you please explain the 0.52 derating factor for me. Does it mean by using an off grid inverter and flooded cell battery I'm loosing almost half of the power generated by my solar panels ?!?!
    Is there any way to improve this value? Perhaps by using certain batteries, high end inverter, etc? Thanks again for your help.

    Mark

    I would argue that a .52 derate is probably too optimistic. In addition to all the system loses, starting with PV's generally only putting out ~80% of name plate on an good day, weather effects, charge controller/battery charging eff/inverter eff etc, another issue that is not spoken about much is power management.

    For example, if you have a "perfect system design" that on an average day puts out 100% of daily draw you would be doing well. Now, however, picture a system that is over sized enough (~30%) to allow the system to fully charge on a day that is not quite average. In that case, that extra 30% would be fine on the less than perfect day, but would be pure waste on the perfect day.

    Another issue is timing. Under the ideal circumstance, your battery would come 100% charged, just as the sun was going off the panel. But once again, assume for a moment that the battery comes 100% 1 hour before the sun comes off the Pv, that 1 hour of available sun is, once again pure waste.

    Finally, load/time shifting can (and does) have a significant effect on power production. For example, if I pump water when ever the P-tank calls for water, I might pump from the battery at night, and by using the battery, I am building in a ~20% loss just by using the battery (since it takes ~20% more power to recharge a battery than the same energy quantity drawn). If I time my water pumping until the sun is powerful enough to power the power the pump, I have gained 20%. But if I wait until the batteries are full, and pump with what would be waste, I gain more efficiency.

    So you can see how living off grid comes with either a significant learning curve as to how use your power efficiently, or a system that is large enough such that there is way more power available (net/net) than one needs, leading to,, more waste.

    In my system, we use ~.6-.8 kwh/day from the batteries, but we can generate ~2 kw. We try to time loads, like charging power tools, pumping water, charging lap tops etc to coincide with excess capacity, but on the good days we often throw away potential after ~1pm. It's just the nature of reality.

    If I was to quantify it all I would suggest that on balance, I am getting a (useful) harvest of maybe 30-40% of name plate.

    Tony
  • Markab
    Markab Registered Users Posts: 18
    Options
    Re: First solar system inquiry

    Thanks BB & Icarus for the replys. I'm going to digest some of this info before moving forward.

    Mark