Can I run without batteries?

System

Hello, this is my first post here. Thanks for a very informative forum.

I'm from South Africa and I'm interested in buying a small tract of land in an off-grid location. I want to run a small machine shop in a shed on this property. This will include running stuff like an electric drill, sander, angle grinder, compressor and a small lathe.

I'll only be using these power tools during daylight hours so the thought has occurred to me to run without batteries altogether and instead drive the loads directly off an inverter connected to the PV panels. I figured that the elimination of batteries would make the system almost maintenance free and significantly reduce the cost. If clouds obscure the sun or if it starts raining then I'll just wait until the sky clears up again before using my tools again. No problem. We are blessed with really abundant sunshine in this part of the world, by the way.

Is this at all possible? I would assume that the amperage would vary wildly throughout the day when connecting an inverter directly to the panels. Would an inverter be able to cope with this or would I need some sort of voltage regulator to go between the panels and the inverter? I'm a true novice at PV systems but I figure that one could use a couple of capacitors to accommodate the starting surge of AC induction motors, right?

Thanks again.

icarus

Re: Can I run without batteries?

Not to be terse,,, but NO!

Do a search on this site, the question has been answered quite often.

I'm sure others with chime in with all the reasons, but put simply, if you add up the amperage draw of your tools and their run times, I suspect that very much of your power will go to waste. For example if the run time of a drill is 25% and you have capacity enough to run that drill, what is going to happen when the drill goes off? As far as I know most inverters need to "see" battery voltage to work properly. The only direct panel operation that I know of is water pumping with pumps connected directly to pump controllers and pumps. I have never heard of any other way to go batteryless.

Also to size a system for the starting current of tools would require a "huge' array. Even if you factor in Capacitor to give a start I think it would be huge. ( I know less than nothing about caps!)

Tony

The fact is if you have grid power available, far and away your cheapest option would be grid tie,, efficiencies run in the +90% range. Battery systems around 50% net/net.

niel

Re: Can I run without batteries?

most times i think you would've been advised to run a generator due to the size of the system you'd need and the costs involved. if you are still going solar, then yes you need the batteries. you could run the solar and if you haven't enough power for your tools due to less sun or more tool use you could then supplement with the generator. any way that you do this it will be a bit of an expense as generating your own power is usually far more expensive than going with a utility. the usual exception seems to be when the utility charges an arm and a leg just to run their power lines to you and solar then isn't so bad.

System2

Re: Can I run without batteries?

Thanks a lot for the detailed replies, Gentlemen. May I ask something else?

OK, so I'm going to take it as a given that I'm going to need batteries. Suppose I install a 2kW array (10 x 200W panels) and suppose none of my power tools come close to a power draw of 2000W (apart from their start-up current, which may be higher, of course), then would I not be able to get by with a minimal number of batteries (like 1 or 2 x 100AH's) when the array does in fact produce more than enough power to drive the tools directly. This way, the batteries will merely assist with starting loads as well as to ensure a stable voltage supply to the inverter instead of being the source for most of the actual power.

Am I still off the mark here?

icarus

Re: Can I run without batteries?

The short answer is yes,,,, and no!

The basic efficiency of a battery/inverter/Pv system is ~50%, any sizing has to take that number into account.

You will have to do a number of calculations to determine the smallest size battery bank you will need to couple to what size PV array to power your expected loads.

As Neil suggests, Generator power is, in all likely hood, going to be cheaper.

Tony

BB.

Re: Can I run without batteries?

Mr. Arnie (I call anyone "Big" Mr.) ,

You are correct... You can do this. Your practical limits is the amount of current the battery can withstand (charging and discharging) without over heating / or sagging voltage output (internal resistance and limits of chemical reactions).

Typically, automotive batteries are designed to be light, cheap, and provide short bursts of large amounts of current to start the engine.

Deep Cycle batteries are designed with heavier, thicker plates (less surface area) so that they can cycle deeper and not have the plates fail. However, they cannot supply as high as current (for same size battery) as an automotive battery.

In your application, a true deep cycle battery will probably last longer than an automotive/RV/Marine type battery.

OK... Lets use some of your numbers as a starting point. 2,000 watts of solar panels. And a continuous load of ~2,000 watts maximum.

Lets assume that your tools are ~230 VAC @ 50Hz. And you are using flooded cell batteries with ~80% efficiency and an AC inverter with 85% efficiency (these are conservative numbers...).

I will use the Trojan Battery Sizing Guide--Good batteries from a good company (PDF download). You would look for batteries and their charts from your local supplier.

Now, you may be assuming your 2,000 watts of solar panels will be producing 2,000 watts from 9am to 3pm... In reality, they will probably only produce 80% of that number on an average day--and the peak production will only be an hour or so on either side of noon. That is one reason you will want the batteries. If you are not using the power for a few moments--the extra power from the solar panels is stored in the batteries for use later. And if you use a bit more power than the panels can provide--then the batteries are there to provide it (instead of the inverter simply shutting down).

And, when you account for all losses (panel derating, charge controller losses, battery losses, inverter losses, etc.), the overall formula will work to something like a derating factor of 0.52 (52% overall efficiency).

Amount of sun for South Africa:

Most areas in South Africa average more than 2 500 hours of sunshine per year, and average solar-radiation levels range between 4.5 and 6.5kWh/m2 in one day. The southern African region, and in fact the whole of Africa, has sunshine all year round. The annual 24-hour global solar radiation average is about 220 W/m2 for South Africa, compared with about 150 W/m2 for parts of the USA, and about 100 W/m2 for Europe and the United Kingdom. This makes South Africa's local resource one of the highest in the world.

Let us assume that you can count on an average of 5 hours of sun per day (year round average?). I am sure we can find better data--but this will work for now. Remember, this is the equivalent of 5 hours of "noon time" sun... But it is really collected over 6-8 hours per day. So, you will be relying on the batteries to collect more energy at noon for use earlier/later in the day.

A steady current into/out of a battery is typically rated as Capacity divided by time in hours. So, a typical rating for a battery is C/20 -- or a 20 hour discharge rate. The batteries do not get overheated from these current levels and are what we normally size an off-grid solar system to use.

In your case, I will try a C/5 (5 hours to 100% discharge a battery). A very heavy current--and probably about the minimum C/X value I would suggest for a workable system without going to some other type of battery (NiCAD, NiMH, or some other type of cell).

At C/5 current levels, for Trojan Storage Batteries (pick 6 volt flooded cell), we are going to loose another ~20% in capacity because of these heavy currents (more losses).

Or, now our solar panel to AC inverter output becomes:

0.52 * 0.80 = 0.42 efficiency.

Now, we can look at your system in two different ways... How much power will 2,000 watts of solar panels produce:

2,000 watts * 5 hours of sun * 0.42 = 4,200 Watt*Hours (or 4.2kWhrs) of electricity.

Or, you can define how much power your tools use, and work backwards. Say you use an average of 1,000 watts 8 hours per day, or:

1,000 watts * 8 hours = 8,000 Watt*hours per day (8 kWhrs per day).

The amount of solar panels needed:

2kW of solar panels/4.2 kWhrs produced * 8 kWhrs needed = 3.8 kW of solar panels

It will be up to you to determine how much power you will use in an average day... Your best bet would be to run the tools through a kWatt*Hour meter (typically your utility meter, or buy a used kWhr meter, or get an industrial kWhr totalizing meter)...

For the sake of argument, lets assume that you will be using 1kW (1,000 watts average 8 hours per day) and need 4.2kW of solar panels.

Next, we need to pick an inverter. Guess you will need a 230 volt 3kW inverter (50 Hz) to start your tools (and it is a True Sine Wave inverter--more expensive, but more efficient when working with electric AC motors).

Generally, I like to tell people that they want to limit their average current from their battery bank to about 100 amps maximum. You can parallel more batteries together to higher amperage--but the copper cables become very thick and expensive.

Using the formula Power=Voltage*Current:

P/V=I=3,000 watts / 48 volt battery bank * 1/85% inv eff = 74 amps (using a 48 volt battery bank).

Again, you can use a 24 volt (or possibly even a 12 volt) battery bank--but I would highly suggest 48 volt bank (if you can find an appropriately rated inverter).

For example a 12 volt 3kW inverter would draw 4x as much current, or ~290 amps--a whole lot of energy.

Also, it will turn out that solar battery charge controllers are current limited (80 amps for the larger ones). They will work at 12 or 48 volts--but at 48 volts, they can manage 4x the number of solar panels vs at 12 volts (hope that was clear).

Now, we have picked the inverter input voltage, and now we know the battery bank voltage (48 volts).

We are going to now jam C/5 charge rate (20% of battery Capacity Rating) into the battery. So, for a 3.8 kW solar panel array:

3,800 watts of solar PV / 60 volts into battery = 63 amps

If we pick 6 volt batteries, we will need 48v battery bank / 6v batt = 8x 6 volt batteries in series.

Each battery will need a C/5 capacity of:

63 amps * 5 = 315 amp*hours at a 5 hour rate

Or (remember, I guessed at a 80% derating for specifying batterys at C/5 vs C/20 rate):

315 AH * 1/0.80 = 394 AH at a 20 hour rate (normal battery rating)

A ~394 amp*hour battery is a "L16" case which will weigh around 50 kg (120 lbs) each--and you will need 8 of them.

You will now need a solar charge controller. Outback makes a FM80 (80 amp) unit which will work very nicely for your application (just need one of them).

So, in summary:

3,800 watts of solar panels
3,000 watt TSW 230 VAC 50Hz 48 volt DC input Inverter (True Sine Wave)
8x ~394 Amp*Hour of L16-P Trojan Deep Cycle Batteries (400 kg)
1x Outback FM-80 solar charge controller

And it will produce:

1,000 Watts over an 8 hour day (assuming 5 hours of sun per day)

Because of the heaving currents, you probably will need lots of distilled water (or clean and filtered rain water) to keep your batteries properly filled (perhaps 1-4 gallons per month?).

OK--now all of the disclaimers... In the above, I have made a whole bunch of assumptions. Your batteries which would last 10 years in normal service may last only a few years in this severe service.

I everywhere I have made, for the most part, pretty conservative assumptions. You may get 40% more power out of your system on during the summer (good weather).

The solar panels are assumed to be Fixed Mounted Panels. If you tilt your panels for summer/winter angles--you may get more power. Or you could put them on a single or dual axis tracker to gain power too.

The above are just initial guesses at sizing so you can get an estimated cost of your system.

Because you are operating outside the normal rules of thumb--it is difficult to guess how your system will actually perform (or how long the batteries will last).

If this sytem is still of interest to you... It would probably be a good idea to build a smaller system first (say 1/4 size) and see how it works for you--make sure you have a good Battery Monitor and an AC kWhr meter so you can check the system's performance and see how well it actually generates power.

If I think about the system more, I may see errors in my assumptions or calculations--so please go through everything I did and please ask if something does not make sense.

And, as always, we suggest conservation first... Whatever you choose, make sure that the tools are the smallest/most power efficient you can find. And if you have choices for alternatives (sky lights / windows instead of electric lighting; solar heated drying chamber instead of electric heater; low pressure water pumping instead of high pressure; etc.)--take them. It is almost always to conserve a watt vs generating a watt.

What you have asked is a tough question--I hope I was up to it.

-Bill

PS: You could possibly size the batteries smaller, possibly even down to 100 amp*hour per battery--but I am not sure I would recommend that--the battery life would probably be terrible. And you would be spending much of your time watching the Battery Monitor to ensure you are not killing you bank after an hour of work...

System2

Re: Can I run without batteries?

Wow! What a great reply, Bill! I am really indebted to you for going to such trouble to help me. I'm going to have to spend a couple of hours digesting the wealth of information that you've provided.

May I offer a few more point to consider. I'll understand if you don't feel like replying; you've helped me quite enough already.

1. I'm not going to be able to afford more than 2kW worth of PV panels. Everything else will have to fit in with that number.
2. I've checked with my local authorities and I can pretty much work on at least 6 full sunshine hours per day and even in excess of 6.5 hours in summer.
3. My tool use will be pretty intermittent and I'm not envisaging even close to 8 hours per day of usage. The typical stuff that I do in my workshop involves as much measuring, welding (with gas),surface finishing, forging, melting etc. as it does the actual use of power tools. I'll probably be more than happy with the ability to run my tools for a total of 3 hours per day.
4. I'll be willing to put up with intermittent interruptions and do other things while these are in progress (like clowds obscuring the panels).

I would assume that the above parameters would enable me to get away with a far less elaborate system, not so?

BB.

Re: Can I run without batteries?

Mr. Arnie,

Check my post count--obviously I probably need to pace myself a bit more that I have been.

I did give you all of the equations/assumptions I have made. If you wish, put the formulas in a spread sheet and you can change them very quickly.

For a quick estimate, just cut all of my numbers by ~1/2 (2 kW of panels instead of 3.8 kW). And up the total kWhr output by ~20% (6hr/5hr of sun = 1.2x).

If your usage is highly variable (some days more than others, or even a couple days of out-of-shop time)--you may be happier with more batteries. They are the "storage" tank of your power system.

At least one person here has been pretty successful running his fabrication business in Hawaii by picking up old forklift batteries and using them for his shop.

Search posts from user "ADAS". He has also been picking up surplus MSW (modified sine wave) Inverters and been pretty happy with them too (you can find them pretty cheap in the US).

Adas' setup is not one I would suggest to everyone--but if you have the time and the inclination (plus sources for the parts)--it has worked well for him.

-Bill

BB.

Re: Can I run without batteries?

Mr. Arnie,

I forgot to discuss your 3 hour a day power usage...

I assume you already are using your electric tools? Either "in town" with utility power, or in "the bush" (?--or is that Australia) with a generator of some sort.

Try to see if you can get a kWhr meter--I used to get them and wire up an power cord into them... Plug one end into the power source (utility/generator) and run all of your loads on the other. And keep track of the meter readings (every day or once a week / 5 or 7 days) to figure out how much power you are really using.

I would recommend this to you even if you are just using a generator. You can quickly see how much you are paying for your electricity (fuel+maintenance costs / kWhrs used).

Many people have a way oversized generator. Its nice to have a 10kW genset, but if you only use 500-1,000 watts of power 99% of the time--you can be using several times the amount of fuel that a smaller 2kW genset would use instead.

A fairly efficient gasoline power small genset (2-5kW) can generate ~ 5-6kWhrs per gallon of fuel or:

5kWhrs per gallon / 3.786 liters per gallon = 1.32 kWhrs per liter
6kWhrs per gallon / 3.786 liters per gallon = 1.58 kWhrs per liter

And I am sure a good quality diesel generator will do better... But unless you measure your fuel usage and kWhrs used--you will not know how good (or how bad) your fuel economy is.

Here is an example of a industrial kWhr meter that I was suggesting.

-Bill

System2

Re: Can I run without batteries?

Thanks a lot, Bill, you've been most kind. In fact I do have a kW/h meter and I'll do exactly as you suggested. I'm still on the grid at the moment, by the way.

icarus

Re: Can I run without batteries?

I lied,,, I said I would write more later,,,,, If I wait long enough Bill will do all the writing for us,,, and better too!:D

Tony