Off-grid project - some technical Qs

[omerk]

Hey guys,

I have searched many forums for answers to my questions and I think this forum will be it

I am designing an portable off-grid hybrid generator (Solar & Wind).
It is supposed to be stationed in the desert (high temps, very wide temp range between day and night, lots of dust).

The configuration is as follows:

- 18 Panels of UniSolar PVL-144 (weight nothing, no glass - highly durable and work well in high temps and lower light conditions(covered in dust)).
- 1 Tangarie 2.25Kw VAWT.
- around 1000AH @ 24V battery bank AGM deep cycle (concorde/trojan).

This is the basic system, and this is where my questions are just rising all the time.

Q. which configuration give more yield?

Should I go with AC coupling configuration? Means I will connect the solar panels to SMA sunnyboy and the wind turbine to Windy Boy and have Sunny Island as mini-grid?

Or should I connect the panels on dc side with sunny island chargers (MPPT) and the turbine on the AC side and have the Sunny ISland manage the grid and charging?

The consumer draws 700W per hour 24 hours a day... I think that the second configuration yields more power at lower light conditions where the SunnyBoy would pass the limit to start generating energy - Is this correct?

Q. Connecting the turbine on dc side to yield power at low wind conditions:

Is there any charge controller that I can connect 0-600DC from the turbine so I can connect it to the dc side(meaning it will sync with sunny island)? Will this mess with the Sunny Island chargers if they won't be synced? Having 2 different chargers not sync on the same battery bank.

Q. How do I determine the threshold where the inverter/charge controller starts producing energy?

Meaning I see a charge controller have "optimal MPPT range 25V-60V", "max input voltage 140VDC", "Max input current 40A", "Max Pv power 1250W" - What is the parameters for starting the harvest the energy from the panels?

Q. Which AGM batteries would be best from your experience?

As you may noticed I chose AGM because they don't spill anything while this system is on the move and they have good characteristic for off-grid applications.

Which company is better out of Trojan and Concorde? Better performance, pricing and guarantee.

Comments and improvements are welcomed!

[peterako]

Hi there,

If you want to make it portable, then leaf out the vawt wind turbine. And spend the money on more panels. Around 800 lbs is not portable plus you will not have the energy that you want.

Also 700 Watt per hour 24 hours a day is a lot for 24V 1000AH. you are below the 70% from the energy whitin 24H. one day of heavy clouds and your system is down.

You dit not inform about any backup generator????

Greetings from Greece

[omerk]

Hi,

Thanks for the quick reply.

Portable meaning it can be towed(Max 2 Tons). - And the turbine mast is folding into half of the height and then turned horizontally to allow the unit to be highly portable...same for the solar - they fold into a third of the deployed size.

I have used the turbine exactly for those days when there is no sun (meaning there is high probability of wind - storm, night etc..) - The array from the solar panels is already too big(by physical size) and I can increase it by 3 more modules - no more than that - it is already a big "sail".

There is a backup generator, however, the client terms is that the system will have an hourly uptime of 70% per year.

About the battery bank I might increase it a little to 1200AH...It is planned to by cycled down to 40% before the generator starts and 60% for the generator to turn off and the system to go on batteries again.

[tallgirl]

What Peterako was saying is that your system's loads are just too big. 700 watts, 24 hours a day, is an incredibly large load. Adding three more solar panels, or 200 more amp-hours isn't going to fix that problem.

You have to reduce the 700 watt load. 18 x 144 watts per panel is 2,592 watts -- assuming the sun is shining directly and perfectly on the panels. 24 hours x 700 watts is 16,800 watt-hours before conversion losses are added in. Assuming 65% efficiency (it will very likely be less for you) for a battery backed system, you'd need 25,850 watt-hours per day from the wind and sun.

[BB.]

- 18 Panels of UniSolar PVL-144 (weight nothing, no glass - highly durable and work well in high temps and lower light conditions(covered in dust)).
- 1 Tangarie 2.25Kw VAWT.
- around 1000AH @ 24V battery bank AGM deep cycle (concorde/trojan).

I understand the low weight of the UniSolar panels--however, they are probably 1/2 the efficiency of standard crystalline panels. Physically, you are going too need 2x the structure to hold the UniSolar panels vs crystalline--Do they really save you anything in the end? Granted, they are less likely to be broken (no glass).

Regarding dust covered operation--both sets of panels produce power based on the amount of solar radiation hitting the cells... Cover either with 50% dust, both will lose 50% of their power output. I would suggest that you setup up 1 each UniSolar vs Crystalline solar and confirm actual structural support requirements and output characteristics.

I would check any wind turbine and how well it is sealed against dust. Wind turbines tend to be less than reliable anyway--and dust will be a killer for both types.

I am not sure I would go with a VAWT at all... Wind turbines need lots of non-turbulent air and, ideally that would work much better if mounted, at the very least, on 60' minimum tower (get into higher altitude prevailing winds). A VAWT mounted near the ground simply will not be in the wind anyay--And VAWT's in general have very poor history of measured power production and reliability.

Concord (as I recall) is one of the few vendors that claim their batteries operate well when discharged to 20% state of charge--Since you need "light weight"--I would check their batteries out closely.

Q. which configuration give more yield?

Should I go with AC coupling configuration? Means I will connect the solar panels to SMA sunnyboy and the wind turbine to Windy Boy and have Sunny Island as mini-grid?

Or should I connect the panels on dc side with sunny island chargers (MPPT) and the turbine on the AC side and have the Sunny ISland manage the grid and charging?

The consumer draws 700W per hour 24 hours a day... I think that the second configuration yields more power at lower light conditions where the SunnyBoy would pass the limit to start generating energy - Is this correct?

You have conversion losses anyway you look at it. If your heart is set the wind turbine -- I would contact Midnite Solar directly. They are in Beta with their new MPPT type solar/wind charge controller ("Classic" model). May offer substantial increase in wind turbine power collection through MPPT function for wind.

700 watts 24x7--Is this for something like a Cell/Sat station in Afghanistan?

Q. Connecting the turbine on dc side to yield power at low wind conditions:

Is there any charge controller that I can connect 0-600DC from the turbine so I can connect it to the dc side(meaning it will sync with sunny island)? Will this mess with the Sunny Island chargers if they won't be synced? Having 2 different chargers not sync on the same battery bank.

Yes, you can put two (or more) unsynced controllers on one battery bank.
I don't know of an MPPT charge controller that can take 0-600 VDC wind turbine output... Midnite may take up to 250 VDC...

Q. How do I determine the threshold where the inverter/charge controller starts producing energy?

Meaning I see a charge controller have "optimal MPPT range 25V-60V", "max input voltage 140VDC", "Max input current 40A", "Max Pv power 1250W" - What is the parameters for starting the harvest the energy from the panels?

Solar panels will produced a voltage > Vmp in fairly dim light... However, realistically, it takes direct sun (something that throws a shadow) or bright overcast to produce useful energy.

The typical MPPT charge controller only has the capability to down convert voltage--So, when you see a range of 25-140 VDC (typically for a battery charge controller)--The minimum Vmp of the panel should be:

• Vmp>= Vbatt-charging + ~2 volts for controller/wiring drop

So, for various battery banks (note, I used ~14.5 volts for AGM, other flooded cell may require 15.-15.5 volts for equalization):

• Vmp >= 14.5 volts + 2 = ~16.5 volts for a "12 volt bank"
• Vmp >= 29 volts + 2 = ~31 volts for a "24 volt bank"
• Vmp >= 58 volts + 2 = ~60 volts for a "48 volt bank"

Q. Which AGM batteries would be best from your experience?

As you may noticed I chose AGM because they don't spill anything while this system is on the move and they have good characteristic for off-grid applications.

Which company is better out of Trojan and Concorde? Better performance, pricing and guarantee.

I would suggest Concord as a good start--they claim the widest rated cycling range (20-80% State of Charge). Also, if I recall correctly, you can really stuff the current down them when needed (say a large genset to recharge back to 80% state of charge).

Comments and improvements are welcomed!

A few observations... Wind Turbines of any type, and VAWT specifically, have very poor history of power generation. An interesting set of data here done in Netherlands of small wind:

Small windpower a scam ? Survey says SO

What ever type wind system you choose--Elevation is critical. A small turbine up high is most likely going to perform much better than a large turbine down low. For example--some rules of thumb for wind distribution:

Ground Drag
The avoidance of ground drag will increase performance dramatically. Up to a considerable height, the least expensive way to increase your power output from a wind turbine is to increase tower height.

A generally recognised 'rule of thumb' is that wind speed increases as the 1/7th power of the height above ground. The following curve illustrates this theoretical increase in wind speed with increasing height above ground:
[see website for chart]
As an example in the use of this curve, if a windspeed of 15 kph were measured at 2 metres above the surface, the windspeed at 20 metres height can be predicted from the curve.

At 2 metres height, the 1/7th power is 1.104, and at 20 metres it is 1.534. Dividing 15 kph by 1.104 and then multiplying by 1.534 yields the predicted windspeed of 20.8 kph at 20 metres.
However, the energy in the wind, and therefore wind generator output, is proportional to the cube of the windspeed. So, in this example, by increasing the tower height from 2 metres to 20 metres increases the wind-turbine output by 2.67 times.
[see website for chart]

Next, if it has to be light, easy to tow, etc... If this is a temporary installation--a fuel driving genset is difficult to beat--or at least should be a standard to compare alternative solutions with... For example:

• Honda eu2000i genset running at 700 watts will give ~5.45 kWhrs per gallon of fuel
• 2 tons of fuel = 4,000 lbs and 6 lbs per gallons (gasoline) = ~670 gallons
• 670 gallons * 5.45 kWhrs/gallon = 3,650 kWhrs of power
• 3,650 kWhrs / 0.700 watt load = 5,200 hours = 217 days

Granted, a Honda euX000i is not the correct geset for this application--just using for fuel calculations (relatively fuel efficient at this load level).

If this is a long term installation (multi-year) then solar and batteries can make economic sense. And then a fixed installation with rigid glass solar panels would seem to make sense too.

But, light and mobile, with low(er) up front costs--an appropriately sized prime mover genset pair is going to be difficult to beat.

I would suggest sizing the system based on solar+fuel genset... And you can add wind as an experiment (site by site, vendor by vendor performance testing). At this point, there are probably only a few reliable wind turbine vendors out there--and I am not sure that any of them are VAWT manufacturers.

-Bill

[stephendv]

Q. which configuration give more yield?

Should I go with AC coupling configuration? Means I will connect the solar panels to SMA sunnyboy and the wind turbine to Windy Boy and have Sunny Island as mini-grid?

AC coupling works out slightly more efficient only if you consume more than 60% of the generated power directly, without it going into the battery. As soon as you start putting that power into the battery you hit the double conversion losses that Bill mentioned.
There's an excel sheet comparing the efficiency of AC vs DC coupling over here:
http://www.casanogaldelasbrujas.com/...d-ac-coupling/

Or should I connect the panels on dc side with sunny island chargers (MPPT) and the turbine on the AC side and have the Sunny ISland manage the grid and charging?

DC charging will give the you best charging efficiency, but I wouldn't bother with the SMA charger, it's max 40A and is laughably expensive compared to other MPPT chargers on the market. Morningstar, outback, xantrex all have MPPT chargers will almost double the capacity for half the price.

The consumer draws 700W per hour 24 hours a day... I think that the second configuration yields more power at lower light conditions where the SunnyBoy would pass the limit to start generating energy - Is this correct?

Not sure what you mean here? Check the speadsheet on the link I posted, you can tweak the variables to compare efficiencies, if this is what you're referring to.

Q. Connecting the turbine on dc side to yield power at low wind conditions:

Is there any charge controller that I can connect 0-600DC from the turbine so I can connect it to the dc side(meaning it will sync with sunny island)? Will this mess with the Sunny Island chargers if they won't be synced? Having 2 different chargers not sync on the same battery bank.

If you connect any DC charging source (non-SMA) or DC load with a sunny island then you need the additional current shunt from SMA, this will let the sunny island know what's going into and out of the battery so that it can keep track of the state of charge.

[omerk]

Hi,

Thanks for all the replies...I'll try to not miss something...

I understand the low weight of the UniSolar panels--however, they are probably 1/2 the efficiency of standard crystalline panels. Physically, you are going too need 2x the structure to hold the UniSolar panels vs crystalline--Do they really save you anything in the end? Granted, they are less likely to be broken (no glass).

High durability is a necessity so I can't have the panels being broken when the truck comes and lifts the structure or from rocks etc... other than that I have seen that Uni-solar being the most efficient out of the amorphous thin film also have triple pass diode which helps it achieve good yields in clouded and if the panel is covered in some section it doesn't render the whole panel useless only the cells which are covered...and at high temp the amorphous technology has the advantage.

I am not sure I would go with a VAWT at all... Wind turbines need lots of non-turbulent air and, ideally that would work much better if mounted, at the very least, on 60' minimum tower (get into higher altitude prevailing winds). A VAWT mounted near the ground simply will not be in the wind anyay--And VAWT's in general have very poor history of measured power production and reliability.

The Vawt is placed at around 6 meters height... the best you can get from portable structure, at this high vawt is better than hawt since we have done testing in the field and the hawt don't switch direction fast enough and they can't harvest the gust wind because of that. About reliability, this is an important issue yet to be tested, we chose a company(Tangarie) that has installed in various climates (Ocean bouy, Antarctica, Northern Africa)...however good point mentioned here.

You have conversion losses anyway you look at it. If your heart is set the wind turbine -- I would contact Midnite Solar directly. They are in Beta with their new MPPT type solar/wind charge controller ("Classic" model). May offer substantial increase in wind turbine power collection through MPPT function for wind.

Thanks! I will check the additional companies, it is just that SMA has very good representative in my country and the rest have mediocre representative who don't understand much of what they are saying.

Is SMA that bad? Or just plain expensive?

700 watts 24x7--Is this for something like a Cell/Sat station in Afghanistan?

Bingo...it planned to be for that application, therefore no fuel around...and can't rely on one energy source (sun)...the structure is moved every few months, it sits on a base like this:



So pricing is less of an issue, I need the best products to have the best solution for all round any weather - that's why I chose Uni-Solar and that is why there is 2 power sources (wind and sun) - there is a backup generator but it should work more than 20% of the time...

After checking with Concorde they have mil-std batteries and have higher capacity than Trojan so I think I will go with their AGMs.

Q. If I leave the SMA Sunny Island as the "manager" of the grid between the battery bank and the device and the generator and have 2 other (un-synced) charge controllers (one from solar, one from wind) all charging the batteries it is ok? It doesn't do anything to the batteries?

If I understood correctly from stephendv all I need is a current shunt from SMA for each charge controller?

[BB.]

Omerk,

From what little I know--SMA should be a good supplier (expensive, but solid designs).

Looking at the 50 meter wind maps of Afghanistan (PDF), only in the western edge of the country do you have significant areas of good or better winds.

I really have a question about your VAWT and 6 meter tall mounting... I looked at the vendor website (www.tangarie.com ?) and did not see any ratings...

The following are giant guesses about your wind installation using data from multiple sites...

The swept areas of the Tangarie are not that large, and depending on which model you are thinking of using--say the middle model Gale 5--Even in the Class 4 or better areas from the Afghan wind map (maybe 10-20% of the country)--call it 7 m/s at 50 meters--or around 4.7 m/s at a 7 meter hub hight for the VAWT would generate less than 2.6 kWhrs per month per sqft:

• 21.52 sqft [for Gale 5] * 2.643 kWhrs per month per sqft (at 4.92 m/s) = 56.9 kWhrs per month
• 56,900 Whrs per month * 1/30 days per month * 1/24 hours per day = 79 watts average (based on HAWT turbine efficiencies)

An near ideal Gale 5 wind turbine at 6 meters in the windiest western region of Afghanistan would generate around 10% of your average load of 700 watts...

To be honest, it hardly seems like it is worth packing a 1,000 lbs of wind turbine and tower for that small amount of power.

Have you actually characterized the Tangarie?

Turbulence near ground level kills any turbine's output. I have not seen any studies that show that VAWT are significantly better than HAWT in turbulance--and in fact, on a swept area calculation Savonius drag type turbines (which the Tangarie appears to be) are even much less efficient than the HAWT bladed estimates I used above (at best only 1/2 as efficient based on sq.units)... So a Gale 5 may produce only 5% of your average needed power. Plus wind is usually highly seasonal--where as solar is more spread out through the year.

On the other hand, Afghanistan is great for solar.

Regarding Charging the AGM's--You probably cannot charge with too much current into the Concord AGMs (you will damage your cabling first--the Concords have been reported to take up a C*4 rate--full charge in 15 minutes?--you need to verify).

The bigger issue with all AGM's is they are very sensitive to over charging (if you over charge, they will vent hydrogen / electrolyte and fail soon after). So, your charge controllers should all have remote battery temperature sensors (typically mounted to the positive post of the battery). As lead acid batteries heat up, their charging voltage drops. Will be very important in the wind ranging temperatures of a high desert.

Another question--your battery bank is 1,000 AH at 24 volts? For a 700 watt 24x7 load that is not very much battery bank. Assuming a maximum discharge of 80%:

• 1,000 AH * 24 volts * 0.80 max discharge * 1/700 watts = 27.4 hours of storage

While you can do that--it will probably require much more genset run time. Wind tends to be highly variable and there will be days when the sun is not out... The longer you can run on battery bank, the more chance you will have to "catch up" with solar (or wind) on the following day or so... But, you are limited in weight--so more batteries are probably out of the question (normally, we recommend ~3 days of no sun and 50% maximum discharge for longer life--but battery life/replacement costs are less of a worry for your operation).

However, if it was a choice between another 1,000 lbs of battery vs a 1,000 lbs of wind generator--I would take the batteries. Each Gale 5 is probably, on average, saving you 5 gallons (~30 lbs) of fuel a month. Get rid of a 1,000 lbs of wind turbine and add 150 gallons of fuel (real rough estimates assuming fuel efficient gas or diesel genset).

Sorry for the meandering reply... Lots of issues.

-Bill

[halfcrazy]

I assume this is a government contract you are competing for? How far along are you? I assume there are still many competitors at this stage if memory serves correctly you would have a complete design and possibly a prototype by the time it is down to the single digits of company's competing?

I will say The standard Hawt probably wont hold up there but there are a couple company's producing a turbine that will but They may be direct competitors of yours not sure.

The Vawt will be an issue they will not hold up and the turbulent wind is worse I would vote for a toughened Hawt any day

[stephendv]

If I understood correctly from stephendv all I need is a current shunt from SMA for each charge controller?

You just need 1 current shunt in total, doesn't matter how many charge controllers you have.
There's nothing wrong with the SMA charge controllers except the price - they're big solidly built units and they integrate seamlessly with the sunny island. To give you an idea of costs, with a 24V battery and 2.6kW of solar, you'd need 3 of them for 120A max (approximate retail price of 2400 euro for 3). You could get 2 x outbacks FM60 for a little over 1000 euros.