Newbie with quite a few questions

steven6282steven6282 Posts: 2Registered Users
Hello everyone,
First I want to apologize, I'm sure some of my questions get asked all the time but I've been doing research on the net for the past few days and I just keep finding things to confuse me more. So I'm hoping if I ask my questions someone can explain them to me in a way that makes sense.

I'm looking at doing some experiments in building a solar system. Right now the goal is simple, to power my PC off grid as much as possible. I have a beefy system, under normal conditions it runs around 250 watts. Under a full load while I'm playing games on it or something it may peak as high as 550 watts (I've gotten these numbers with a Kill a watt unit).

Now I know that means I'm using between 2 to 5 amps on a 110 volt wall supply. So my first question deals with inverters!

If I had a solar system, and was using say a 12 volt battery bank (I really don't think I'll be able to get by with this low of voltage but I'm using it for an example for now), am I correct in assuming this means my amp draw on the battery bank will be ~ 20 - 41 amps? (This is of course not taking into account efficiency loss in the inverter).

If that is correct then I need something more like a 48volt battery bank because I don't see any batteries out there that would not be drained in a matter of minutes at this rate. That brings to mind another thought on batteries. I see a lot of batteries rated at say 20 hour 100 AH. From what I've read I gather that means the battery will last for 20 hours at a 5 amp draw. But I'm confused by the 20 hours part. What is the purpose of this being there? The 100 AH is enough to already tell me that it will last for 20 hours at a 5 amp hour drain (or 10 hours at 10 amps, or 40 hours at 2.5 amps.. etc). Does this rating have some other meaning that I'm not aware of? What I'm worried is it may be a max discharge rate. If that is the case even with a 48 volt bank of batteries like that, my computer would be pulling to much drain when it's at max load (550 watts / 48 volts = 11.45 amps).

Then the voltage of the battery bank brings me to my next question. Charging them with solar power. I'm looking and since this is all intended to be a side project for me I want to build the solar panels myself. The cells readily available on ebay are generally either 3x6 cells (1.8w, .5 volt) or 6x6 cells (3.9w,.5 volt). If I built a panel using 36 of the 6x6 cells that would give me a panel with ~135 watts and 18 volts. Everything I read shows using the same voltage in the battery bank as the solar output to charge the batteries. If I built 3 panels in series they would give me 54 volts and 405 watts. Could I charge a 48 volt battery bank with that? This of course is peak efficiency on them so what happens on cloudy days when they might be operating at below 50% efficiency (27 volts 202.5 watts). Would they still charge the batteries some?

Concerning the wiring on the batteries, how does that work? If I have 2 post batteries, I'm essentially going to have two connection points for charging, and discharging. So that means the wires running from my batteries to the inverter would be in circuit with the wires coming from the solar array. Would the inverter not bleed power from the solar array and prevent it all from being used to charge the batteries? And would this not cause an overload in voltage on the inverter getting power from the batteries and power from the solar array at the same time?

Assuming all of my numbers and figuring is accurate, it's going to take me a ridiculous amount of batteries and solar panels to power one computer, how the heck do people power an entire house with solar power? If the batteries were 100 AH I set up banks of 48 volts, I'm going to be draining one bank in 8 to 20 hours depending on the activity on the PC. So realistically if I wanted to do it right I would need 3 banks in parallel to have enough power to possibly last a day or two of cloud cover (If I'm lucky). Oh and that is completely not taking into account the recommendation to only drain the batter to 50% at the lowest before charging. So in that case I would have to double it! And then the panels to charge a bank that large without taking a week to charge it. Using the panels I stated above, assuming the were capable of operating at max efficiency for 10 hours a day (not likely), they'd be able to charge 75 AH. If I had 6 banks of 48 volts at 100 AH, that means it would take 4 days of perfect sunlight to charge them fully from 50% charge, assuming 0 drain on the batteries for those 4 days.

Surely I've got something wrong, I can't imagine my single computer requiring this much, even if it is a bit more power hungry than an average system.... The costs of setting up a system like this are ridiculous, even doing a lot of things myself. The batteries alone would be thousands of dollars...

Comments

  • CariboocootCariboocoot Posts: 17,615Banned ✭✭
    Re: Newbie with quite a few questions

    Welcome to the forum.

    Let's play with numbers!

    You have a known load you want to power, and its maximum is 550 Watts. That means first of all that the inverter has to be capable of supplying at least that much. This you need to know as it eliminates choices like the Morningstar 300, which has a very low self-consumption (power needed to run the inverter itself).

    The next number that comes up is the Watt hours for a given day. This can also be measured with a K-A-W meter under typical usage conditions. That's the one that has the most bearing on the battery bank size. For example:

    550 Watts * 8 hours a day = 4.4 kW hours.

    But the Watts may average less than that and the time may be more or less.

    250 Watts * 8 hours a day = 2.0 kW hours, 250 Watts * 6 hours = 1.5 kW hours, et cetera.

    Now, leaving out the inverter's power demands for a moment if you divide those consumption numbers by nominal system Voltage you get an approximate number of Amp hours used. In this example anywhere from 125 Amp hours to 367 Amp hours on a 12 Volt system. Since you don't want to go below 50% SOC on most batteries that number needs to be multiplied by at least 2 to get the battery size. Pretty big 12 Volt system.

    So it's time to read this: http://forum.solar-electric.com/showthread.php?15989-Battery-System-Voltages-and-equivalent-power

    As you can see, the more stored power required the more advantageous higher system Voltage is.

    So how do we power whole houses off grid? First thing we do is reduce power consumption. My whole cabin uses <3 kW hours per day. Surprisingly, the computer equipment is one of the biggest power users because it is on for hours straight. Even though it uses 1/3 as much power as the refrigerator when running, the refrigerator only runs 1/3 of the time.

    But it is not impossible to supply large amounts of power, just expensive. You could in fact have a 1200 Amp hours 12 Volt battery bank or even larger. It just gets to be a pain to do it (large wires, large current, multiple charge controllers).

    If you look at that big 4.4 kW hour number on 24 Volts you get 183 Amp hours and on 48 Volts you get 92 Amp hours. So 220 Amp hours @ 48 Volts would do the full power with just one string of eight golf cart batteries, one charge controller capable of 22+ Amps, and about 1400 Watts of panel.

    How are we doing so far?
  • BB.BB. Posts: 27,882Super Moderators, Administrators admin
    Re: Newbie with quite a few questions

    Welcome to the forum Steven!
    steven6282 wrote: »
    Hello everyone,
    First I want to apologize, I'm sure some of my questions get asked all the time but I've been doing research on the net for the past few days and I just keep finding things to confuse me more. So I'm hoping if I ask my questions someone can explain them to me in a way that makes sense.
    If we did not answer questions a second time, we may as well just shut down the forum--And close all schools while we are at it. :D
    I'm looking at doing some experiments in building a solar system. Right now the goal is simple, to power my PC off grid as much as possible. I have a beefy system, under normal conditions it runs around 250 watts. Under a full load while I'm playing games on it or something it may peak as high as 550 watts (I've gotten these numbers with a Kill a watt unit).

    You will now see why we say your first three steps for going off grid solar are: Conserve, conserve, conserve.

    For a gamer or somebody using a computer for modeling/video processing/etc... They will suck the power down like nobody's business. Computers (home, office, server farms, etc.) are supposed to draw 30% of the total US power grid (who knows if those numbers are accurate--or my memory for that matter :p)
    Now I know that means I'm using between 2 to 5 amps on a 110 volt wall supply. So my first question deals with inverters!

    If I had a solar system, and was using say a 12 volt battery bank (I really don't think I'll be able to get by with this low of voltage but I'm using it for an example for now), am I correct in assuming this means my amp draw on the battery bank will be ~ 20 - 41 amps? (This is of course not taking into account efficiency loss in the inverter).

    Actually, it is really even a bit worse when you take losses and "real life" into account... for example, 550 Watts on an AC inverter and 12 volt battery bank, using wost case (nominal) engineering math:
    • 550 Watts * 1/0.85 inverter eff * 1/10.5 inverter battery cutoff = 61.6 Amps from "low battery" and some voltage drop

    And such a circuit (wiring, breakers) would be (per NEC):
    • 61.6 amps * 1.25 NEC derating = 77 Amps minimum rated breaker+wiring (round up to 80 amps)
    If that is correct then I need something more like a 48volt battery bank because I don't see any batteries out there that would not be drained in a matter of minutes at this rate.

    Not exactly... Power = Voltage * Current ... If you have 4x the voltage, you have 1/4 the current. However, the weight (and cost) of the battery bank would be the same. And notice that AH is missing "voltage" for energy storage -- So by itself, AH is not useful for our needs.

    For example. 4 x 100 AH @ 12 volt batteries
    • 4 x 12 v batteries * 100 AH in parallel = 400 AH * 12 volts = 4,800 WH of storage
    • 4 x 12 v batteries * 100 AH in sereis = 100 AH * 48 volts = 4,800 WH of storage (the same)

    More or less, to choose the battery bank voltage, we look at your power needs. Very roughly, 100 amps is a lot of current (and gets even larger with safety factors/worst case voltage/voltage drop)--So, very roughly, the maximum system wattage at each voltage would be:
    • 12 volts * 100 amps = 1,200 watts max
    • 24 volts * 100 amps = 2,400 watts max
    • 48 volts * 100 amps = 4,800 watts max

    Now, a big problem is voltage drop in the wiring. For 12 volts you have only about 0.5-1.0 volts of "headroom" for wiring losses. At 24 volts you have 1-2 volts, and 48 volts you have 2-4 volts you can support for voltage drop.... Each means that higher voltage battery banks are easier because you can use smaller gauge wiring and send the power longer distances than low voltage battery banks.
    That brings to mind another thought on batteries. I see a lot of batteries rated at say 20 hour 100 AH. From what I've read I gather that means the battery will last for 20 hours at a 5 amp draw. But I'm confused by the 20 hours part. What is the purpose of this being there? The 100 AH is enough to already tell me that it will last for 20 hours at a 5 amp hour drain (or 10 hours at 10 amps, or 40 hours at 2.5 amps.. etc). Does this rating have some other meaning that I'm not aware of? What I'm worried is it may be a max discharge rate. If that is the case even with a 48 volt bank of batteries like that, my computer would be pulling to much drain when it's at max load (550 watts / 48 volts = 11.45 amps).

    Because batteries are not "perfect"... The more current you draw from a battery, the less "apparent capacity" it has... For example a T105 6 Volt battery looks like:

    185 @ 5 hour discharge rate

    225 @ 20 hour discharge rate

    250 @ 100 hour discharge rate


    For a "typical" off grid home/cabin design, for various reasons, the 20 Hour Rate is most useful for us, and we use our Rules of Thumbs around that 20 hour rate / battery capacity.

    If you have very heavy, but short loads, such as water pumping to a pressure tank, perhaps the 5 hour rate is better.

    If you have a small remote computer and you want 4-6 days of storage, then the 100 Hour rate may be most useful.
    Then the voltage of the battery bank brings me to my next question. Charging them with solar power. I'm looking and since this is all intended to be a side project for me I want to build the solar panels myself. The cells readily available on ebay are generally either 3x6 cells (1.8w, .5 volt) or 6x6 cells (3.9w,.5 volt). If I built a panel using 36 of the 6x6 cells that would give me a panel with ~135 watts and 18 volts.

    Don't do it... Not worth the time or the money. Plus the panels will not last in weather and can be a fire hazard (unless you build them "correctly" which is not easy--plus the EBay cells are usually rejects because they are not sutable for use in solar arrays--they have current defects which can cause fires too):

    Panel Fire Question


    Yes, you can make panels--But there is no good reason to do it (other than personal edification). They are not appropriate for bolting to the roof of a home/cabin.
    Everything I read shows using the same voltage in the battery bank as the solar output to charge the batteries. If I built 3 panels in series they would give me 54 volts and 405 watts. Could I charge a 48 volt battery bank with that? This of course is peak efficiency on them so what happens on cloudy days when they might be operating at below 50% efficiency (27 volts 202.5 watts). Would they still charge the batteries some?

    No, not with those numbers... More or less, solar panel's working Vmp (voltage maximum power) is variable with cell temperature, and can easily drop by 20% on hot days... So, the Vmp-array for a 48 volt battery bank (which requires 60 volts to properly charge) would need a Vmp-array minimum of ~72 VDC.
    Concerning the wiring on the batteries, how does that work? If I have 2 post batteries, I'm essentially going to have two connection points for charging, and discharging. So that means the wires running from my batteries to the inverter would be in circuit with the wires coming from the solar array. Would the inverter not bleed power from the solar array and prevent it all from being used to charge the batteries? And would this not cause an overload in voltage on the inverter getting power from the batteries and power from the solar array at the same time?

    Yep, you need to design the system correctly (for safety and correct operation). More or less, look at your car's (or boat's) DC electrical system. That is what happens with an off grid power system... The alternator is a charging source, just like your solar array is. The battery supplies energy when the charging sources are not enough (motor off, sun set) and are recharged when there is available charging current.
    Assuming all of my numbers and figuring is accurate, it's going to take me a ridiculous amount of batteries and solar panels to power one computer, how the heck do people power an entire house with solar power? If the batteries were 100 AH I set up banks of 48 volts, I'm going to be draining one bank in 8 to 20 hours depending on the activity on the PC.

    Yep, if not worse. And you don't want to cycle a lead acid battery bank less than 50% state of charge during normal operations--Banks deeply cycled will have much shorter life. And taking a bank below ~20% SOC, can kill a bank (most of us have "murdered" our first set of batteries).
    So realistically if I wanted to do it right I would need 3 banks in parallel to have enough power to possibly last a day or two of cloud cover (If I'm lucky). Oh and that is completely not taking into account the recommendation to only drain the batter to 50% at the lowest before charging. So in that case I would have to double it! And then the panels to charge a bank that large without taking a week to charge it. Using the panels I stated above, assuming the were capable of operating at max efficiency for 10 hours a day (not likely), they'd be able to charge 75 AH. If I had 6 banks of 48 volts at 100 AH, that means it would take 4 days of perfect sunlight to charge them fully from 50% charge, assuming 0 drain on the batteries for those 4 days.

    Probably--I have not looked in detail at your math.
    Surely I've got something wrong, I can't imagine my single computer requiring this much, even if it is a bit more power hungry than an average system.... The costs of setting up a system like this are ridiculous, even doing a lot of things myself. The batteries alone would be thousands of dollars...

    Yep... Conservation, conservation, conservation. Then off grid solar. Desk top computers, refrigerators, and other loads that run 12-24 hours per day are some of the worst power consumers. A microwave or well pump running 20 minutes a day is much less of an issue for off grid power design (just need a large AC inverter to get enough power to the device for that short period of time).

    Give us some "real numbers", and we can go through the math to give you a starting point. It is usually easier to see/understand the first time through if we design "your system" rather than hand wave all of the choices to be made for every sized (small to large) system possible.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • steven6282steven6282 Posts: 2Registered Users
    Re: Newbie with quite a few questions

    Thanks a lot for the helpful answers! Sorry it's taken me so long to reply again but it's given me a bit to think about. And I think for now I'm going to have to scale back my solar experiments to something quite a bit smaller lol. I just can't afford to spend 5000 + dollars on a system just to power a single computer. Computer right now is about 25 dollars a month on my electric bill, if I spent 5000 dollars to take it off grid, it would take me over 16 1/2 years to break even! Not an economically sound choice for something that is ultimately intended to be a side project / experiment for me.

    I checked my electric bill and my house is currently averaging about 3kw an HOUR! I can't imagine the amount of cutting back we'd have to do in order to get to 3kw a day. I actually don't even think it would be possible in the summer and maintain a livable environment. Our house is old and not well insulated, plus our AC is somewhat inadequate and old too. Means during the summer the AC runs 24 x 7 and fights a losing battle during the day. If I was going to spend 5000 dollars on something right now, it'd probably be a new AC unit :)

    Right now I'm thinking about doing a smaller system just for experimenting / learning purposes with a single 12 volt (or 2 6 volt) batteries and one solar panel capable of charging it. I can probably power a low powered PC like my NAS (uses around 25 watts of power in it's normal state, up to around 60 watts if I'm actively streaming something from it) on something like that.
  • CariboocootCariboocoot Posts: 17,615Banned ✭✭
    Re: Newbie with quite a few questions

    You've already learned the most important thing; conservation gives better payback than generation. :D

    It is not difficult to design an experimenting system around constraints because you have the grid available if the power runs out. It's when you must depend on the batteries that things get tricky - and expensive.
  • westbranchwestbranch Posts: 5,087Solar Expert ✭✭✭✭
    Re: Newbie with quite a few questions

    Well, if it was my house and I had the choice of PV vs Insulation, it would be to re-insulate or add insulation to reduce the A/C runtimes before your A/C unit goes down due to 100% duty time. Start with the attic and try for R40....

    hth
     
    KID #51B  4s 140W to 24V 900Ah C&D AGM
    CL#29032 FW 2126/ 2073/ 2133 175A E-Panel WBjr, 3 x 4s 140W to 24V 900Ah C&D AGM 
    Cotek ST1500W 24V Inverter,OmniCharge 3024,
    2 x Cisco WRT54GL i/c DD-WRT Rtr & Bridge,
    Eu3/2/1000i Gens, 1680W & E-Panel/WBjr to come, CL #647 asleep
    West Chilcotin, BC, Canada
  • BB.BB. Posts: 27,882Super Moderators, Administrators admin
    Re: Newbie with quite a few questions

    Yep--I stripped the sheet rock off my walls and stuffed in fiberglass insulation plus installed double pane vinyl low-e windows (paid a contractor for the insulation, did the windows myself--they were existing double hung wood windows and were reasonably easy over all). Later installed a 97% efficient condensing central furnace (make sure you sump pump/pit is "reliable"--furnace no like to swim :blush:).

    Made the home a lot nicer to live in and quieter too (late 1930's stick frame home with redwood siding--no insulation at all, except some blown in attic insulation from a couple decades ago.

    Home was always too hot or too cold... Now, sort of cool in the summer (insulation helped) and only a ~440-$50 per month winter natural gas bill (excluding hot water/cooking at ~$20 per month). Of course, we live in a very temperate climate. Your mileage may vary.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
Sign In or Register to comment.