Greetings from Haiti!

Barnstormer

Hello!

I've just arrived in Haiti to spend a year doing facility direction/maintenance on a hospital here in the mountains of Haiti. One of the things I have inherited is a reasonably big solar system.

We are totally off grid here, with 2 separate systems powering about 4 buildings. One system is 40 panels feeding 36 6v batteries, the other system is 24 panels feeding 24 batteries (I think- I haven't counted it all)
It's all 48v, I've got 4 total Xantrex inverters, 4 Trace C40 charge controllers, and some misc other stuff.

Previous people have had even less knowledge than I when it comes to setting up the values in the charge controllers- the bulk and float voltage values, especially. I am sure the system can be much more efficient than it is currently.

So, I'm looking for advice on the setup of the charge controllers, especially. What the best bulk and float values should be, etc.

I'm sure I'll be picking the brains of everyone here- thanks in advance for all the help!

KeithWHare

Re: Greetings from Haiti!

Barnstormer wrote: »

Hello!

I've just arrived in Haiti to spend a year doing facility direction/maintenance on a hospital here in the mountains of Haiti. One of the things I have inherited is a reasonably big solar system.

Which mountains? The mountains of Haiti cover a good bit of the country.

We are totally off grid here, with 2 separate systems powering about 4 buildings. One system is 40 panels feeding 36 6v batteries, the other system is 24 panels feeding 24 batteries (I think- I haven't counted it all)
It's all 48v, I've got 4 total Xantrex inverters, 4 Trace C40 charge controllers, and some misc other stuff.

When you get a chance, start posting more details about battery, panel and inverter models and post pictures of the hookups, etc.

Also look at the wiring and grounding -- wire sizes, etc.

I can't make 36 6-volt batteries work out to a 48 volt battery bank.

The Trace C40s are pulse width modulated (PWM) controllers, so they won't be quite as efficient as a MPPT (maximum power point tracking) charge controllers.

Previous people have had even less knowledge than I when it comes to setting up the values in the charge controllers- the bulk and float voltage values, especially. I am sure the system can be much more efficient than it is currently.

So, I'm looking for advice on the setup of the charge controllers, especially. What the best bulk and float values should be, etc.

What information can you dig up about power usage, peak loads, etc.?

I'm sure I'll be picking the brains of everyone here- thanks in advance for all the help!

Provide as much information as you can. I'm pretty sure you will get a lot of response, both from people like me who have learned enough to sometimes ask the correct questions, and from people who actually know the answers.

(Here's a thread on an installation I did in Haiti, just east of Port au Prince http://forum.solar-electric.com/showthread.php?t=9127)

Keith

Barnstormer

Re: Greetings from Haiti!

I'm smack in the middle. Thomassique is the town we are in. Here, if anyone is interested.

Sorry, it's 32 batteries in the one pack, 24 in the other. So far, my investigation says things are built pretty well- wire sizes are all quite sufficient, and it all seems to be functional at the moment- if not ideal.

I've got 4x Xantrex SW5548 inverters, Trojan L16RE-A Batteries, 4x Trace C40 Charge controllers, and a mix of Kyocera KC120-1 and KC130TM panels (not sure the #s of each, pretty tough to get to and see at the moment)


I checked the settings on the charge controllers, they were all over the place- it looks like someone just randomly turned the bulk and float settings. I set them all back to a guess as to where they would be better- but I'm not sure.

I'm plenty comfortable with all the rest of the systems here (normal 120 AC system, plumbing, etc.., but the solar is pretty new to me.)

Luckily, I do have the manuals for the inverters and the charge controllers- but I don't know what the best bulk and float values should be for a system like this.

BB.

Re: Greetings from Haiti!

Barnstormer,

A question for you--How comfortable are you with 1960's electricity (current, voltage, batteries, etc.)?

Mostly, your system will evolve around the battery bank... Keep them happy, and they will give you a long life. If they are treated roughly, the batteries may last a couple of years or less--and send your costs for power way up.

A couple of good FAQ's to start with:

Deep Cycle Battery FAQ
www.batteryfaq.org

For checking out and diagnosing the system... I would suggest a good DMM (digital multimeter) and an AC/DC type Clamp Amp Meter (this one is $100 or so). They are not that complex to use--a 1/2 day of using them and you will be able to diagnose most problems with the systems.

For long term operation by less technical folks (and to make things easier for you too)--I would suggest a Battery Monitor too (or from Victron Energy). This is the equivalent of adding a gas gauge to your car's fuel tank.

And, of course, some sort of hydrometer + thermometer to measure the battery bank's specific gravity (get a couple, they are easy to break).

We can argue the best voltage and current to pump into your battery banks--but measuring/monitoring the state of charge with a hydrometer+Battery Monitor will tell you what you have to do to keep the batteries "happy".

-Bill

Barnstormer

Re: Greetings from Haiti!

The two separate systems consume different amounts of power. At the moment the staff house (which has the 32 battery pack) is consuming 6 amps, and it seems that it peaks about 15A. There are 2 full size AC fridges, a couple microwaves, and a number of ceiling fans, lights, phone chargers, computers, satellite internet system, and who knows what else running.

At the moment, (1:30pm) with a pretty sunny day today, I've got 21 and 23 amps on the two charge controllers, and the "battery actual" on the inverters says 50.4v. The charge controllers say 50.0v (why the difference?). I do have thermocouples in the packs tied to the inverters.


I'm far from a licensed electrician, but I understand batteries and electricity- I'm plenty comfortable working with grid based systems, it's the solar production end of things that I'm not very familiar with.

Barnstormer

Re: Greetings from Haiti!

Also, at the moment (and for the last couple months) there has been no generator backup for the system. A new 25Kw Cummins/Onan is on the way, but delivery times are a mystery around here. SO, the system has not been equalized for at least that long- and very likely never. Once the generator arrives and I get it online, I'll equalize the packs, which I'm sure they desperately need.

Another issue is that I know that the batteries have been drawn very low many times since the generator failed. Overall, I'm sure many parts of the system have not been treated very well- maintaining things isn't the strong suit down here.


This is definitely a "make do with what you've got" kind of situation- replacing things isn't really an option, unless there's a large failure.

BB.

Re: Greetings from Haiti!

Knowing your peak usage such as Amps and Watts is important--but we also need to know the time (hours) too.

For example--We have two loads. One is a 1,000 watt microwave and the other is a laptop computer that draws around 30 watts average.

Say they use the microwave 30 minutes (1/2 hour) per day. And the laptop for 12 hours a day:

• 1,000 watts * 0.5 hours = 500 watt*hours
• 30 watts * 12 hours = 360 watt*hours per day

So, even though the microwave would appear to the "an energy hog"--Because it runs such a short time, the small/power efficient laptop uses a similar amount of power on a 24 hours per day setup.

You can talk about Amp and Amp*Hours vs Watts and Watt*Hours--but those of us with bad memory are going to keep asking at what voltage (48 volts for both your systems?).

So--You need to get "control" of your loads and figure out how much is being used, and probably, how much you will need to cut back on/conserve until the genset arrives and installed.

For some rough rules of thumbs... The battery bank is roughly 80% efficient... If you pull 2,000 WH (2 kWH) out, you will need to return ~2,500 WH (2.5 kWH) back into the battery bank to fully recharge it...

Amp*Hour wise--You need to return, at least, the amount of AH you take from the batteries as a load (100 AH out, you need to return around 100-120 AH back in). Note that you discharge around 50 volts and recharge around 59 volts--So that is the majority of your "battery loses" come from (when you equalize, then you are pushing Amp*Hours into the battery, but not really adding anything to the capacity--so Equalization is pretty much all losses--you only do it, typically, once every 1-6 months depending on need. Too much equalization will kill a battery bank too).

OK, talking about your battery bank... At 50 volts when charging--that is quite a low voltage and indicates that your battery bank is severely in need of recharging.

Assuming the batteries are around 80F, you want the resting voltage to be around 12.7 volts (50.8v for 48 volt bank), and around 12 volts (48 volts) would be 50% discharged (again, after resting a few hours).

For charging, you want to be around 14.4 volts (57.6 volts), and if you ever need to float the battery bank (no loads for a period of time), you can set the float to 13.7v (54.8v).

Basically, you really would like to cycle the batteries around 25% of their capacity every day, maybe every so often (bad weather, heavy loads) down to 50%. And you never want to cycle them below 20% state of charge (80% state of discharge)... If you go below 20% state of charge (11.58 volts / 46.2 volts resting), you run the risk of permanently damaging on or more cells in the battery bank.

So--At this point, get your batteries back above 80% State of Charge--Ideally above 90% state of charge a couple times a week. You will probably have to go on a real energy diet (no microwaves, no fans, no electric tea kettles, etc.).

Also, you should take your DMM and DC Clamp Amp meter and start looking for bad cells and strings not properly sharing current... Look for batteries/cells that are too low of voltage (shorted cells) or too high of voltage (open cells/boiled dry/etc.).

Use the current meter to see that each string is accepting charging current and supplying current... Strings that are not working properly may have bad cable connections or bad batteries in the string. Try to isolate the bad batteries and, perhaps, take the good batteries and rebuild a string. (It is better to remove bad batteries and have a smaller bank with known good batteries--it will have better function overall).

-Bill

Barnstormer

Re: Greetings from Haiti!

BB. wrote: »

OK, talking about your battery bank... At 50 volts when charging--that is quite a low voltage and indicates that your battery bank is severely in need of recharging.

Assuming the batteries are around 80F, you want the resting voltage to be around 12.7 volts (50.8v for 48 volt bank), and around 12 volts (48 volts) would be 50% discharged (again, after resting a few hours).

For charging, you want to be around 14.4 volts (57.6 volts), and if you ever need to float the battery bank (no loads for a period of time), you can set the float to 13.7v (54.8v).


-Bill

Excellent. This is the info that I needed first.
I think the problem is 2 fold- One, the charge controllers were set to charge at about 50.5v and float at about 49. So, nothing has been charged well based on that.

Second, there doesn't seem to be any understanding of conservation (basically a huge problem that effects almost everything here), so it's going to be very difficult to convince/teach people to use less. The general idea is that if power is coming out of the socket, use it! I discovered the "normal" use of the LVCO was to reset it 5 or so times every night for the 10 additional minutes of power- no concept that it might be bad for anything. I'm sure the technical purists here on NAWS would faint watching how people treat the system here.
I'm going to mostly concentrate on optimizing the system and making and storing as much power as I can- and then trying to control access to the system so it can't be messed with as much as possible.

I've set the charge controllers to the above settings- hopefully that will start to make a difference in overall charge state. The batteries are full of water, the connections have been checked and re-greased with anti-corrosive, and now I've got keys to the locks on the equipment rooms- things are moving towards the better!

I'll try over the next few days to get an idea of the actual consumption, although at some level it doesn't matter, as I have very little pull to try to get folks to use less.

Thanks for the help so far!

BB.

Re: Greetings from Haiti!

To a degree, let the battery banks tell you if they are getting sufficiently charged...

Watch the electrolyte levels (and, of course, fill with distilled or well filtered rain water (captured after the first rains rinse off the roof), and never well water).

If your banks are not using any water--Need more charging, less loads.

If the banks are in danger of exposing the plates after 30 days--then you probably are over charging (may never be an issue for your systems... :roll:).

And, don't fill the cells to the top when beginning a heavy charge--It is possible for the cells to warm up/fill a bit with gas, and push electrolyte out the tops of the cells.

For normal charging--you will see a little bubbling... You do not want to see a lot every day for hours at a time--that may indicate overcharging.

If all you have is a DVM/DMM--when things are charging, go through and measure each battery terminal set... Ideally, they should all be within 1/20th (0.05 volts--assuming 6 volt batteries) of a volt of each other. If you have any batteries that are more than 0.5 volt difference--You have a major problem with that battery/string...

Also, if you do not have a clamp meter... if there is a, for example 2' section of cable, in each parallel battery string... You can use that like a current shunt (resistor). Set the meter to 2.000 or 200mV scale (hopefully, yours can read down to 0.000 or 1/1,000th of a volt/millivolt scale...

For each series string of batteries... Measure the voltage drop (when under heavy load or heavy charging) across the same cable in each series string. Some may read around 0.010 volts, others may read 0.030 volts, and other may read near 0.001 volts... The "high voltage" reading indicates strings taking a lot of current--And the low voltage indicates that string is not participating--and you need to figure out for both...

If you have high current when charging--that string may have a shorted cell. If you have low current when charging, that string may have an open cell or open cable/bolted connection somewhere.

And for discharging--strings that are not providing power may have shorted or open cells/bad cables/connections, etc....

Also, check for things like hot electrical connections (poor/corroded connections) or hot batteries (shorted cell, too much current through string)...

Get a hydrometer (lay flat on soft surface--with edges to prevent rolling off), do not store upright in can/jar--it will fall and break). Monitor the specific gravity of a couple cells to watch state of charge. And check all cells once a month or so to see that they are "balanced" (specific gravity ranges across all cells of 0.030 or greater indicates equalization is needed or other battery issues).

There is quite a bit of stuff you can watch for even if you do not have much in the way of tools.

If they are resetting the inverters several times a night to get a little extra power--Typically those are set to 42 VDC for cutoff... That is 10.5 volts (for a 12 volt battery) and represents a really "dead" battery bank. Doing this can kill a bank in months or even weeks... The "weak" cells actually go through zero volts and start getting "reverse charged"--that typically destroys those cells and that battery is no longer any good.

Perhaps you can work with people in charge to understand how expensive the electricity is... For example--Assume your battery bank lasts two years, and you get 5kWhrs per day of useful power. And say a replacement bank costs $3,000:

• $3,000 / (5kWH per day * 2 years * 365 days) = 0.82 per kWH

If the battery bank lasts 1 year because of bad practices--Then it is going to double the cost of their power...

So--They can go back and look at--for example, say they use an electric tea kettle and use 1kWH per day to heat the water (1 hour use per day)--That is the equivalent of $0.80 to $1.60 "worth of power"... Was that worth it--Or would a kerosene/diesel cook stove have been a better/less expensive energy source.

Good Luck!
-Bill

Barnstormer

Re: Greetings from Haiti!

All good advice. I've got a clamp-on ammeter, I'll start checking and fliddling in the next few days.

Here are some pics of some parts of the system

KeithWHare

Re: Greetings from Haiti!

Barnstormer wrote: »

I'm smack in the middle. Thomassique is the town we are in. Here, if anyone is interested.

OK, looks like you are 15-20 miles east of Hinche. My younger brother, Mark Hare, is an agronomist working with MPP in Papay, a couple of mile North of Hinche. The road to Papay goes to the north off of 308 (at least that is what Google calls it), about 1.25 miles from where 308 starts just north of the bridge on the north side of Hinche.

Mark may still able to find the copy of "Photovoltaics: Design and Installation Manual" that I left with him a year and a half ago. If you manage to track him down and if he can find the book, it might be a useful read for you.

Keith Hare

BB.

Re: Greetings from Haiti!

Quite a nice power station you have there guy!

Given that everything has a life too it (batteries, and electronics)--Is there a "emergency fund" that is being kept for ongoing maintenance and repairs?

As you know, installing is just part of the ongoing costs. If they have to start saving "extra moneys" because the battery banks are getting beaten up--That may catch a bit of attention from the owners/operators of the system.

-Bill

Barnstormer

Re: Greetings from Haiti!

Wow, what a difference a day makes. Just a couple days of production with the bulk and float values set correctly has been very good for the state of affairs here, power wise. Things are charging well, and with a bit of forced conservation (it's good to be able to lock the access to the equipment!), we are working our way up to better battery health.