Solar Street Lighting - Can Metal Halides be used with solar??

[rurokenshi]

Hi,
I am new here, i am an engineer trainee and i deal with sustainable solutions.
I have a question regarding Solar Street Lighting

I heard from someone that while considering solar, the "Running Current" of the solar panels must be twice that of the "running current" of the lamps

I was given this as a reason to not use Metal Halides & High Pressure Sodium Vapour lamps for solar applications

Can someone throw some light on this topic for me?

Thanks

[Mike90250]

You were told BS.

You can use any light source with solar.

If you are completely off-grid, you must harvest and store enough power for 3 gloomy winter days, and accomplish recharge in 1 short sunny winter day.

Using a low power light source, LED, CFL, Metal Halide, High Pressure Sodium Vapor and even incandescent are all feasible, you just need to provide power.

Your best bet is the most usable lumens per watt, at the temperatures your environment will be. Some of the gas discharge lights take a long time to start at low temps.

Then figure your winter power usage for 3 days, in watt hours, double that, to get the proper battery bank size, and then you need enough solar PV to recharge in 1 sunny day. Depending on your winter day/night ratio, and sun angle, you could need 2 - 8 x your bulb wattage, you have to run the numbers for your location. Check the PVWatts.org site for some calculators, USA & global.

[Sunking]

Are you sure you are an engineering student?

Any lighting system can be powered by solar. Question is can you afford it and willing to pay 10 times more for electricity provided by solar for x amount of light. The most efficient lighting systems in order are:

High and Low Pressure Sodium
Metal Halide
T5 & T8 Florescent Tubes
CFL
Mercury Vapor
LED
Incandescent

Here is a chart

[Mike90250]

Here is a chart

As fast as LED's are developing, the chart is outdated, LED spots are now at 100 lumens / watt

http://www.ledlight.com/9-watt-spot-led-light.aspx

Fro lighting , LED's are tricky, they are very directional, unlike other bulbs, which are fairly omni-directional. They don't like heat either.

[Sunking]

Mike for LED's the manufactures are still cheating and fudging the numbers. They do not count the ballast or driver power, and they only measure a single pulse of a cold LED in the focal spot of the light 1 meter away.

If you use the NIL standard they fail miserable and no better than an incandescent. Let's not even discuss the color rendering index of LED's. Put it too you this way, the federal government does not allow LED's in any building because they waste too much energy and the poor light quality they produce.

For home interior illumination nothing is more efficient and better quality light than a T5 Florescent Tube. They produce 110 Lumens per watt including the ballast power with a CRI of close to 100 (near perfect color) at 1 meter 360 degrees around the tube in all directions.

T5, T8, and some CFL's are the only lights you will see in government buildings and all new construction both private and public interior lighting. Exceptions would be Emergency lighting and Egress lighting.

[rurokenshi]

Yeah, well i was told by another experienced engineering graduate (Masters) tat MH & HPSV's are not compatible with solar..and also he was a lighting specialist.. Even if i was a phD guy i would have trusted his judgement!!!

[Sunking]

i was told by another experienced engineering graduate (Masters) tat MH & HPSV's are not compatible with solar

Like I said you can power anything you want with solar. The question is can you afford it by paying 10 to 30 times more (assuming this is a battery system) than just buying it from the electric utility.

I have no idea how many lighting fixtures, wattage, how many hours per day, and the location you are talking about, but watt-hours are watt-hours no matter what the source is.

All you have to do is run the numbers. Let's take an example. So let's pretend you want to install a single Redi-Light. These are the small street lights you would find in a residential area, not the large ones on highways. Redi-Light use a 150 watt Metal Halide. It is located in Omaha NB and operates on a photo cell switch. So worse case is the month of December and January with 16 hour nights.

Ok first you determine the power requirements in watt hours. 150 watts x 16 hours = 2400 watt hours or 2.4 Kwh. Since we are talking about a battery system the best efficiency we can get is 66% so we need to account for that. So a solar panel needs to generate 2400/.66 = 3.64 Kwh is what the solar panel must generate in winter.

Ok the solar panel wattage can now be determines We know it has to generate 3.64 Kwh per day, so all we have to do is factor out the time element to find the wattage the panel must be. In Omaha in December and January receives 3.2 Sun Hours. So now we have the time element. Solar panel wattage = Kwh / Sun Hours = 3.64 Kwh / 3.2 hours = 1.13 Kw or we need a solar panel array of 1130 watts to get the job done.

Ok now time to tackle the battery size. To maximize battery life we never want to discharge more than 20% in a given day. This will give us a few cloudy days and extend the battery life up to 5 years before it needs replaced. So we need a battery capacity of 5 days @ 3.64 Kwh so 5 x 3.64 = 18.2 Kwh. Ok batteries are sized in Amp Hours. To find the right battery we first select the nominal operating voltage. Let's use a 24 volt battery. To to find the Amp hours all we have to do is factor out the voltage from the Kwh. So 18, 200 wh / 24 volts = 758 Amp Hour @ 24 volts.

Charge controller is straight forward take the solar panel wattage and divide it by the battery voltage. So 1130 watts / 24 = 47 amps.

Inverter size is twice the connected load of 150 watts x 2 = 300 watts.

Ok we are done with the calculations we need

1130 watt minimum solar panel, round up to 1200 so we can use 6 200 watt panels.
50 Amp MPPT charge controller
Battery we can go with 750 AH at 24 volts
300 watt inverter.

Now for the fun and educational part, cost estimate. This does not include the cabinet to house the 2000 pounds of batteries, labor, shipping, or misc materials, just the electrical components to make it work.
Solar Panel $3 per watt = $3 x 1200 = $3600
Battery is $140 per Kwh = $140 x 18.2 = $2548
50 Amp MPPT Charge Controller = $450
300 watt Inverter = $300
Total = $6898, lets just say $7000

Cost benefit. The system will produce 2.4 Kwh per day, 73.2 Kwh per month, 878.4 Kwh per year, and 4392 Kwh in 5 years when the battery needs replaced. So your first 5 year Kwh cost are $7000 / 4392 Kwh = $1.60 per Kwh for the first 5 years of operation.

Electric rates in Omaha NB are $0.096 per Kwh. So to go solar would require you to pay 17.39 times more then buying from the utility for the first 5 years. After 5 years time to replace the battery of about $3000 lowering the next 5 years to $0.68 per Kwh.

So is that a good deal or not?

[moonrathna]

Hi king...

Your calculation are perfect i have few questions,

Why you are taking 66% for panel sizing??

And For the 24V system the 200W panel max voltage is 27V approx, Is that enough??

You have done the calculation for street lighting, then how the 6 panels going to be mounted in the pole??

In fact we are the streetlighting manufacturer and we are successfully finished few projects with 70w metal halide solar street lighting. But our system, what we design so far is for 2 sunny days only.

Could you please suggest some manufacturer for battery, inverters and charge controllers, since we face some issues in charging while we put for testing

the Voc from the panel was 20V (180W panel) once connected to charge controller it reduced to 12V minimum slightly above the battery voltage, does this replicates the battery voltage?? Or this is the actual charging voltage?? If not how to measure the actual charging voltage??

Last question, whats ur suggestion on panel (poly or mono)..

Thanks in advance for your time and reply.

Regards,

Ratna

[the captain]

in addition to the above, a 150watt MH requires a little more power for the "wasted" power through the ballast. if it uses a magnetic core and coil ballast it should be about 20 watts or so or if it's a digital ballast, maybe about 5 extra watts. so even more capacity is needed. after reading the other post, it's demoralizing just how expensive it would be just to power such a small thing.