A very basic 12 volt PV system for emergency lighting.

[adam2]

Several recent posts have refered to the instalation of a very basic, 12 volt only battery/pv system to provide very basic lighting in case of grid failure.

I am pleased to give a suggested design for a very simple basic system, together with instalation instructions.

Many on these forums already have relatively large or sophisticated systems, my suggestions are not aimed at these experts, but at those without electrical knowledge who require a basic, user installed system, that costs only a few hundred pounds and can be installed in a day.

SHOPPING LIST AND APPROX PRICES
PV module, size according to finance but probably at least 55 watt, bigger is better. allow from £150 to £300.
Deep cycle battery, 12 volt about 70/80 A/H bigger is better within reason, allow about £50.
Solar charge controller, rated at twice the output of the PV module so as to allow for the future addition of a second module.
Five 12 volt light fittings*, five light switches, allow about £50 in total
An in line vehicle type fuse holder fitted with a 10 amp fuse, and some spares, a couple of battery terminals, about £5 the lot
A 50m roll of 1.5mm twin and earth cable (the earth core is not needed, but twin cable is now so unusuall that twin and earth is probably cheaper)
cost £22

INSTALATION INSTRUCTIONS.

Find a suitable location for the battery, away from children or pets and not exposed to extremes of temperature, a kitchen cupboard may be suitable.
Mount the charge controller near the battery but where it is on view.

Connect the lamps, via switches to the output of the charge controller (the correct terminals to use will be marked on the charge controller)
Run the cable from the charge controller to the first light switch, cut the cable and connect both the red/brown cores to ONE terminal of the switch.
Connect both the blue wires together in a terminal block inside the switch box, run the cable on to the second and subsequent switchs. (in a larger home it may be worth running an additional cable from the last light switch, back to the charge controller, so as to form a ring and reduce voltage drop)
Now connect the lamps to the switches, fit the light fitting to the wall or ceiling as required, and run a cable from the light to the switch, at the switch, connect the blue core to the terminal block with the other blue wires and connect the brown wire to the vacant terminal of the switch.

Now connect the battery to the charge controller, (the correct terminals will be marked on the controller) negative battery terminal to negative on solar charge controller, battery positive to in line fuse holder, and then on to controller positive.

Now test, should work fine.

Having wired the lamps to the battery and controller, and tested, now mount the PV module outside where it gets plenty of sun.
Run a cable from the PV module to the correct terminals on the solar charge controller, connect brown to positive and blue to negative.
Ensure that the "charging" lamp on the controller comes on.

Job done!

*Light fittings should be flourescent for good light output, or LED for lowest power use, a wide variety are available. Avoid incandescent lamps except perhaps for lights used only briefly.
Avoid flourescent lamps of more than 13 watts, since they may overtax such a small system and cause voltage drop in the wiring.
I suggested five lamps as an arbitary number, up to 10/12 could be installed provided they are not all turned on at once.

The total load in use at any one time should not exceed about 5amps/60 watts, in order to avoid excessive voltage drop in the wiring.

In the darkest winter weather it should be possible to use one lamp all evening, together with limited use of the others.

Performance and usefullness could be enhanced by fitting more modules, a larger battery, an inverter,power outlets,and more than a single lighting circuit, this would however add to cost and complexity.

For a large or complex system an approved electrician should be consulted, however a small basic instalation is well within most peoples abilities (no claim is made regarding part pee compliance)

(edited to fix ? £ problem caused by change to site hosting)

[emordnilap]

That is really cool! I'll print out the post and study it. How can we thank you?

[Joules]

Yeah, thanks Adam. Sometimes we non-sparky-types feel inadequate on this forum... well I do anyway.

[hardworkinghippy]

Adam,

Fantastic post - really useful !!

I'm no expert but I live off-grid and I've a few photos to go with your post showing batteries, controllers, wiring, connectors etc. in this set in flickr.

http://www.flickr.com/photos/hardworkin ... 363481598/

Roll your mouse over the small photos to get a text explaining what's in the photo.

[JohnB]

Is 1.5mm? cable big enough? I've got an 80 watt PV module with a 10 amp charge controller to allow for an extra module, and was advised to use 2.5mm? cable. My installation instructions has a table of recommended cable sizes that vary with the length of the cable run. I can dig it out if it helps. My cable run is probably a lot shorter than you could manage in a house.

I use LED lights in the van, and they're great in a small space with a low ceiling, but I haven't tried them in one of those strange house things most of you live in!

[hardworkinghippy]

Here's a table for calculating the right thickness of cable for length of run and DC voltage.

http://www.unlimited-power.co.uk/cable_sizing.html

[adam2]

Is 1.5mm? cable big enough? I've got an 80 watt PV module with a 10 amp charge controller to allow for an extra module, and was advised to use 2.5mm? cable. My installation instructions has a table of recommended cable sizes that vary with the length of the cable run. I can dig it out if it helps. My cable run is probably a lot shorter than you could manage in a house.

I use LED lights in the van, and they're great in a small space with a low ceiling, but I haven't tried them in one of those strange house things most of you live in!

1.5mm should be fine in practice, 2.5mm would be better but adds to costs, and wont easily fit the cheaper light switchs or light fittings.
I did suggest, in larger premises, running a second cable from the last light switch back to the charge controller, this gives an effective cable size of about 3.00mm.

Remember that although 5 or more lights will be installed, it is unlikely that more than two or three would be regularly used at the same time.
Therefore the load current in unlkely to regularly exceed 3 amps (and will often be less) 1.5mm if run in a ring is good for about 20M (to the far point of the ring, that is, i.e. a total distance out and back of 40M), or in a radial about 10M. Detailed calculation is not appropriate because we dont know exactly what the load will be, and also the load will be spread along the circuit, not concentrated at the far end.
There would be no danger in turning on every light at once since the installation is protected by a 10 amp fuse and 1.5mm cable is good for more than 10 amps. (the lights woul be rather dim though)

The cable from the PV module is unlikely under most conditions to carry more than 4 amps (for an 80 watt PV module in real world conditions), again 1.5 mm cable will be fine, though 2,5 might be better.

As a very rough guide, avoiding calulations, one may say.
"map out the cable routes in advance, keeping them as short as feasible, if you find that a single 50M roll of cable is insuficient, then this shows that you have longer than typical runs, and should therefore consider 2.5mm cable"

Most simple instalations in a small house wont use more than 50M of cable, we are talking of a basic system for a small low powered lamp in each main room (or perhaps a lamp in every room, but only 5 watts or less in toilets etc)

[JohnB]

The cable from the PV module is unlikely under most conditions to carry more than 4 amps (for an 80 watt PV module in real world conditions), again 1.5 mm cable will be fine, though 2,5 might be better.

I've seen my Steca charge controller show 4.3 amps coming from my 60 watt module mounted flat on my van roof. It was either today or yesterday, and may have gone higher, but I just happened to spot it. It's a 10 amp controller so I can add another module wired in parallel. In total I have 4m of 2.5mm cable from the module to the battery, with the charge controller roughly in the middle.

adam2 - The Steca manual advises an external temperature sensor close to the battery, if the charge controller is mounted where there could be more than a 4 degree temperature difference between battery and controller. I've fitted one as I get bigger variations than that. Would you advise one?

[adam2]

The cable from the PV module is unlikely under most conditions to carry more than 4 amps (for an 80 watt PV module in real world conditions), again 1.5 mm cable will be fine, though 2,5 might be better.

I've seen my Steca charge controller show 4.3 amps coming from my 60 watt module mounted flat on my van roof. It was either today or yesterday, and may have gone higher, but I just happened to spot it. It's a 10 amp controller so I can add another module wired in parallel. In total I have 4m of 2.5mm cable from the module to the battery, with the charge controller roughly in the middle.

adam2 - The Steca manual advises an external temperature sensor close to the battery, if the charge controller is mounted where there could be more than a 4 degree temperature difference between battery and controller. I've fitted one as I get bigger variations than that. Would you advise one?

4.3 amps from a 60 watt module is impressive, though maybe not that often achieved, 2.5 mm cable would certainly reduce voltage drop and give slightly better charging, whether its worth it on a small basic emergency lighting system is debatable (unless the cable is more than about 5M)

As to the use of an external battery temperature sensor, again for the sort of simple basic system proposed it may be a needless complication.
The very cheapest charge controllers dont allways have provision for an external sensor.
In the type of system disscussed I would forsee the battery in a kitchen cupboard, with the controller above on a wall, or perhaps the battery in an understair cupboard, with the controller nearby. Under such circumstances, large temperature differences are not that likely.

[mikepepler]

And here's a tip for choosing a solar panel if you're in rented property: measure your windows!

I've got a couple of 18W panels that fit in the windows in my office room (still letting enough light in for working at the computer). Crystalline panels are obviously smaller for a given power output.

Also, remember that even a small solar PV panel is better than none. I have a second system that only has a 10W panel and 7Ah battery - it's non-emergency use is to run an LED night light for our pet birds. However, it in sunlight it has plenty of power to charge a mobile phone, run a radio of charge batteries, and would run LED lights for several hours.

[adam2]

And here's a tip for choosing a solar panel if you're in rented property: measure your windows!

I've got a couple of 18W panels that fit in the windows in my office room (still letting enough light in for working at the computer). Crystalline panels are obviously smaller for a given power output.

Also, remember that even a small solar PV panel is better than none. I have a second system that only has a 10W panel and 7Ah battery - it's non-emergency use is to run an LED night light for our pet birds. However, it in sunlight it has plenty of power to charge a mobile phone, run a radio of charge batteries, and would run LED lights for several hours.

Very true, the size system I proposed is probably about the minimum to light a flourescent light for some hours every night in the depths of winter.

However a smaller sytem such as proposed above could still be very useful, either for low power LED lighting or for limited use of higher powered flourescent lighting (saves lighting a tilley lamp when light is only wanted briefly)

As a rough guide, in UK winter conditions one may expect about an hour of "peak sun equivalent" per day on average.
Therefore a 10 watt module should produce on average about 10 watt/hours a day, this would light an 8 watt flourescent lamp for about an hour, or a 1 watt LED for about 8 hours a night

A system such as I proposed with say, a 60 watt module, would produce in winter about 60 watt hours a day, this would light an 8 watt lamp in the main living room for about 5 hours a night, and allow brief use of lamps in toilets, bathrooms, bedrooms etc.

[JohnB]

I've got a spreadsheet where you can enter your power requirements and the number of days of autonomy you want. It works out the required size of batteries and PV modules, and shows results for each month of the year. It's based on PV modules mounted flat on the roof, so would need adapting for correctly angled modules, although they would produce more power.

It's based on starting with a fully charged battery, and the power needed to keep it charged. So it's worth remembering that it's no good installing your system during a power cut, as it could take a long time to initially charge the battery.

[adam2]

I've got a spreadsheet where you can enter your power requirements and the number of days of autonomy you want. It works out the required size of batteries and PV modules, and shows results for each month of the year. It's based on PV modules mounted flat on the roof, so would need adapting for correctly angled modules, although they would produce more power.

It's based on starting with a fully charged battery, and the power needed to keep it charged. So it's worth remembering that it's no good installing your system during a power cut, as it could take a long time to initially charge the battery.

The use of spreadsheets or other calculations is the correct way to proceed for larger or more complex systems, or indeed for small ones when a given load must be supplied reliably (for example for radio beacons on which lives may depend)

However for the sort of small simple domestic emergency lighting system proposed, spreedsheets or other calculations are a bit OTT.
In the type of system I propose, the single PV module will probably be sized by deciding what can be afforded/whats cheap on fleabay/what will fit in the available space.
The battery size would in practice, probably be about 70/80 A/H because that size is sold everywhere at moderate prices, unless of course the user already has a suitable battery, in which case that is what gets used (even if too big or too small in theory)

Having installed the system, it will undoubtably work, and will be a great improvement over sitting in the dark!
The exact loads that can be supplied, will be determined by observing the charge controller, rather than by calculation.
Those with funds, expertise and interest could no doubt design a larger system, with the loads listed, and the required battery and PV sizes carefuly calculated.

My remarks are aimed at those without any prior experience who wish to install a small simple system themselves.
Many without prior experience would find calulations a bit daunting.

[emordnilap]

adam2, I read through your instructions and it's all clear except for one small part:

(in a larger home it may be worth running an additional cable from the last light switch, back to the charge controller, so as to form a ring and reduce voltage drop)

Would you mind explaining that part a little more? TIA

[adam2]

adam2, I read through your instructions and it's all clear except for one small part:

(in a larger home it may be worth running an additional cable from the last light switch, back to the charge controller, so as to form a ring and reduce voltage drop)

Would you mind explaining that part a little more? TIA

The simplest way of wiring this sort of circuit, is to run a twin* cable from the power source, which is the charge controller, to the first light switch.
A second length of cable is then run from the first switch to the second, and again to all subsequent switchs.
Therefore every switch except the last one, will have three cables to it;
1) power from source or previos switch
2) switched power out to lamp
3) power onwards to the next switch

The last switch would have only two cables;
1)power in from previos switch
2)switched power out to lamp
My suggestion was to run a third cable from the last switch (as if feeding another switch) run this additional cable back to the charge controller and connect it into the same terminals as the cable to the first switch.

Therefore the current will have two routes from the charge controller to the light swtiches, one route to switch 1 and a second route to the last switch.
This has the same effect as useing a cable of about twice the size and substantialy reduces voltage drop.
Excesive voltage drop will make the lamps dim, and in extreme cases might prevent flourescent lamps from starting.

Detailed calculation is not appropiate since we dont know what size lamps will be used, nor how far along the circuit they will be.

Others have suggested useing larger cable to reduce voltage drop, this has its merits but adds to costs and may require joints since most cheap light switches and light fittings wont accept the larger cable.