Off Grid PV Solar Array Sizing

∫[(Amperage surplus when sun shine strongly) x (time when sun shines strongly)] ≥
∫[(amperage deficit when sun shines weakly) x (time when sun shines weakly)].

Re charging a battery fully drained out would need higher charging voltage. also please take care of efficiency of recharge of the battery bank.10days outage is bad and battery bank would be costlier than panels. Is there a possibility with grid connectivity for the solar unit with no battery bank or very minimal bank of few hours ?

The site is a remote gas field, unmaned, cost of batteries is not a ctiteria.

Adding a gas generating set circumvents the technical riddle but the very purpose of getting clean energy for the station dies on spot, any one can question if a costly generator ought to be there why at all use a solar system! and who will start the generator when it will be needed.

Rig the remote start generator to a relay actuated by the charge controller when the voltage drops to a certain point. It is much easier than it sounds my friend.

I do agree that auto starting of the generator set would not be a serious issue, but the Client wants a stand alone pv system and that is the limitation.

Any way thanks a lot for the good proposal.

I am sure you meant "power in watts" and not "voltage" when you wrote "Re charging a battery fully drained out would need higher charging voltage."!!!

Remember that in Ohm's Law that V x I = P Watts. For a good reminder, look here:- http://www.the12volt.com/ohm/ohmslaw.asp

When charging a LA battery that is discharged, the voltage at the start of charging will usually be relatively low due to the current drawn, this depends upon the maximum amount of current that can be drawn from the source.....

But a large voltage difference (as you imply) will cause huge currents to flow and could bring the battery to boiling!!!

As the LA battery charges, the voltage will rise, and the current may actually drop as the terminal voltage approaches the maximum voltage of the charging source.

Its only the "difference" between the battery's voltage and the charging source voltage that actually allows a current to flow.

Eventually (assuming a quality charger and not some cheap car charger), the output voltage of the charger will either equalise with the battery at some safe voltage and next to no current flows (except for any losses that may occur) or the charger switches off the charge, or it goes into a so called "trickle charge mode".

Trickle charge mode is dangerous unless the charge level is carefully set up to be exactly equal or less than the "voltage loss" of the battery. That is the trickle only partially replaces the charge loss....

If its higher than the losses, the battery will eventually be overcharged and start gassing (explosive gas mixture of Oxygen and Hydrogen), lose water and have a reduced life span.

By the way, unless you (or the OP) are using leisure type batteries, running the battery for long periods under 12.6 volts will promote sulfation and a short life span. A non leisure battery is really slightly incompatible to solar charging for that reason alone, though many use them.....for a time anyway! They appear so cheap!!

Add a generator to the system. You'll need it to equalize the batteries anyways.

You may Keep battery for say 2 days storage and have a generator to charge the battery, if you have long outages, which may be once or twice in a year. Batteries for 10 days will cost a lot. Consider environment pollution of making batteries for 10days storage. Compared to that a small generator should be better. There are many renewable energy sites which do have a generator to take care of situation when sun/wind/battery are not able to provide the energy required. Generator capacity can be just 50% more in terms of Kw compared to average Kw demand . So that it can charge the battery as well as take the load. As soon as battery is charged or solar power is available Generator gets off .

In addition to the 10 day requirement, you also need to account for darkness of night. Neither seems to be a major issue. Your best option, depending on location would be a tracker system to track the light intensity to ensure maximium solar power output, and multiple batter banks and a computer controlled switching system to rotate which battery banks are suppling the DC/AC inverters to provide usable power in times of need.

The drawback to the tracker systems is the need for routine maintenance, so if that is an issue you would be forced to use a ground mounted system, this requires less maintenance.

Another option would be a smaller solar array designed to only keep the batteries charged and larger battery banks to provide power via inverters.

Either way, this could be a costly project. The up side is that it would be "Green".

If you provide long/lat location along with total electrical load in watts and voltage requirements, I will post a line diagram of a few options for review/comment.

Well the tracker system is only advisable for large systems and you are aware of the limititations.

The only load would be instrumentation and SCADA at 24 volts DC so inverters are not there.

http://www.premiumpower.com/ I'd look at these guys,(or this type of technology) avoid lead acid. They can't handle more than around 2,000 cycles. The flow battery will cycle for decades. You wont have to supersize your array to bring back a full charge in one solar day.

Dear A.A. Khi, The standard factors based on the average daily usage (in Watt hours)of all your appliances, is, as a rule of thumb, to cycle the battery bank 20% per day, thus giving four days without input, the input is, allowing for in-efficiencies in Inverter, cabling, battery cycling, etc, equal to the years usage, averaged per day, with perhaps a tiny bit for safety. Should you require longer periods of reserve, you will need to add the original 20% per day to the size of the battery bank (in Watt/hours). Say you use an average of 1000W/hrs/day, a small load like that you could use a 12 volt battery bank, divide the 1000 by 12 = 83.3 amp/hrs/day, so times 5 will require a 417 A/hr(@C10) battery bank. The last 20% stays in the battery so as not to damage it. If u need 5 days without input, then 500 a/hrs, 6 requires 583.33, 10 days, 917a/hrs, - and as batteries usually go up in 100 a/hr sizes, 10 days would require 6 * 1000 amp (2 volt) cells. If you were in Australia, app. $3000; You will need to use Tubular positive cells, as most deep cycle batteries are damaged if you go below 50% DoD whereas Tubulars are fine at 80% DoD. If you don't use Tubulars you will need generator back up and you will have to have a Generator Centric system, with a largeish Genset, and quite sophisticated control equipment, If you use Tubulars, you can have a Battery Centric system, (and only if you want, a smallish genset, - which only turns on when the batteries are below 80%, and which charges only up to half or 3/4,) as the Solar will do the finish charging, equalising etc, - much more efficient. If your load is much larger you may consider if there is enough Wind for a wind generator as often they are complementary with Solar, - ie windy when stormy and grey. Cheers, Geoff.

Advice by Lookfar is excellent.

A small generator reduces overall cost.Even gas can be stored in Cylinders if it is truly stand alone Even a wind mill can be good.Many times wind and solar are complimentary. Wind has the advantage of getting the batteries fully charged to its peak, when the wind is good. Solar can only do a steady charge. The battery back emf build up prevents faster charge,unless charging voltages are made higher.Yes only tubular batteries be used,which can be cycled to 80% Depth of Discharge and can take overcharge better,to charge to full,unlike flat plate batteries.

Excellent points.

Suggestion:

• Use Gel batteries; they can be kept in fully discharge condition for several days.
• Plan battery backup in such a way they divided in to "10 Banks"; one bank is used per day. You can use auto switching based on depth of discharge.
• At the end of 10th no sunshine day, suppose that you have 10 good sunshine days, then if solar panels have 10% more charging capacity than one bank; it charge all banks in 10 days thus ready for a ten day no sun shine days. If you want all battery bank to be ready in five days, then your solar panel should have 120% capacity and so on. Develop control circuitry in such a way solar panels charge any one bank at time.

All the best!

Dear A.A.Khi

(1) Prior to completing the PV array sizing, pls. calculate battery capacity, depending on load current rating, type of battery, depth of discharge ( generally 70% DOD ), min. operating temp., rate of discharge ( depending on duration for which battery has to support load, for more than 4 days C100 details need to be used ), design margin & aging factor.

(2) While sizing PV capacity, sunshine is considered to be available for 5 hrs. average duration ( at least this is how it is done in India, though this will vary with geographic loacation ). If battery is sized for 5 days backup then PV can be sized to charge in 2 SUNSHINE DAYS ( read 10 hrs. by above analogy ). Thus PV capacity should be sufficient to feed load for 48 hrs. ( 10 hrs. during sunshine + 38 hrs. discharging ) plus recharge battery to full state of charge, from 70% DOD ( affectively 70% battery capacity ), in 10 hrs.

If such a prolonged non-sunshine duration ( 5 days ) is not likely to occur at site, frequently, then PV can be sized to charge in 3 or even more, SUNSHINE days also, resulting in lower PV capacity.

Availability of sunshine hours is completely site specific and therefore it is difficult to make a judgement. In India, battery bank in solar pv plants at offshore oil platforms is sized for 5 non-sunshine days, even though 5 non-sunshine days never really occur. Battery bank for gas pipeline application, on land, is sized for 3 non-sunshine days.

This posting is based on my experience of 25 years with manufacturing, supplying & installing, solar charge controllers for offshore oil platforms & 2 years for land based.