2 hours ago, Czauto said:

I'm starting to change my perspective on the whole alternative energy thing.

Welcome Czauto ... it has been quite a journey. I learned that lesson a very long time ago. "Wants" vs "Needs" and their price tag.

As you I think everyone who comes here for help, must first be talked out of solar, before we talk them into it - if that makes sense.

New angle. Since I found out that I can use my existing Eskom meter, no minimum fee, with SMA grid tie with no feedback, 1500w panel max, that is definitely now in MY future. And re-using what I have makes a lot of sense, and it can be done legally where I am.

And the batts I have, existing systems stays in place and for another 2-8 years I can power my fridges and lights and things off the batts at night.

My next journey has just begun, just need the ship to now arrive also.

To be brutally honest here , well done guys!!

This is by FAR the best points that has been made in a long time! @Energy may I ask that this thread become a "sticky" as this will definitely help any newcomer who is looking into Solar.

Paul

 
 

Well... uuhhhm... while I agree with most of you guys... *eyelid twitch*.... *Drool*.... this is so SHINY....

I'm seriously considering something like this one. I just have to do the math and check if it makes sense. I mean, we already know it barely makes sense at retail price with post-taxed money, but in my case this will actually enable me to do more work in this domain, so it's more of an active investment than just a battery.

Edit: Oh, prices. Around 5700 Euro, or ZAR 85k, for 5kwh, 2000 cycles at 80% DoD. About par with what you'd pay for an LG Chem or Pylontech battery. Also, after 2000 cycles it isn't dead, it still has more than half its capacity left.

Just want to add - for the city users - and NOT for the "new shiny toys" route either nor the 5/4/3 guys.

 

The solar journey must start with reducing your needs to a level that makes solar more sense - reduce then solarize.

The problem however comes in, for solar justification, that if you do it like that, that once you have spent the monies to reduce the needs electricity requirement (that costs a pretty penny too), that the usage is now so low that to go solar does not make that amount of sense anymore.

I.e. say your usage was R4k per month and after doing the effort and spending the monies on more energy efficient devices, solar heating / gas, switching things off etc etc etc, you are now down to say R1000 pm MOSTLY spent on night loads, as you are at work 5 days of the week and no-one is home. 

To reduce that +-R1000 pm will cost more in batteries than you will ever save on in not using Eskom for even a constant 200w load is a LOT in battery costs, and batteries will get replaced.

Versus if you are working from home or want to power your offices, i.e. when the sun shines you need the power, then grid tied makes 100% sense, if done legally.

Few last points:

• Never take a loan to buy a solar system. The sums then do not add up at all.
• When you calculate your ROI on the solar investment, factor in the replacement / repair costs. Replacing a inverter or battery bank is not small change.
• Make sure your insurance covers the solar system. Talk to you insurer / broker, get it in writing, insist on it.

I'm soooooo glad we're all agreeing to something at last. 

Going full off-grid in areas with Eskom supply is stupid (except if you're a trustfund kid with too much money). Areas without Eskom or where Eskom bills the Shite out you, yes, I agree going solar and you probably would have your usage on the minimum anyway seeing that a Gennie would have been doing the dirty work anyway,

I was planning to add 6 more panels to my system but why? I cannot even use everything I generate now to the full. If I want to push my efficiency I have to hammer everything I've got the whole day while there's sunshine. Aikona! We use appliances that's needed during peak sunshine and that's where it stops. The extra electricity needed goes through my prepaid meter so no extra fee's on that. I use average 200Kwh per month (just over R400) from the grid. If I would want to invest even more to spend even less than that on Prepaid I would be stupid. There will always be tinkering and wanting more and better but let's face it. It's the same with cars. You always want faster, newer, bigger wheels....always! don't even argue. That just built into any human mind. But hey! Why should I drive a R400k sportscar that can do 300kmh when I can drive my R95k 330d that's 12 years old on a road that's only permitting 120kmh? I get there the same time as you with your half a mill shiny car. No vehicle is an investment and I starting to think the same with alternative energy. Some may need it, some may want it, but think twice if you're looking at that shiny new Mercedes or Audi (or Blue inverter for that matter) If you haven't got a lot of money laying around in your piggy bank, a full on solar system might not be the investment you think........

6 minutes ago, Czauto said:

I'm soooooo glad we're all agreeing to something at last. 

In fact, I just did the math on that nice Lithium bank and my goodness... for the average guy that works out around R10/kwh. That's Diesel Generator territory. If you can subtract the tax (for a business setup) it gets better, and if you consider that Li-Ion degrades better and will likely last way past the 2000 cycles listed (at which point it's still got 70% of its original capacity), I can beat that number down to around R5/kwh. It would take 8 years at 10% increases to get there at present city rates (around R2/kwh).

If you can get that kind of battery at half the price... then it starts to make sense, and that is simply not an option for the average residential customer. If you can get hold of one of these directly from the manufacturer at cost... that would be the closest.

The good news, however, is we're halfway there. Us computer people generally care about "orders", ie things being 10 times (an order), 100 times (two orders) etc faster or slower. A simply constant (k=2) is good news to us! :-)

11 minutes ago, Czauto said:

I'm soooooo glad we're all agreeing to something at last. 

Going full off-grid in areas with Eskom supply is stupid (except if you're a trustfund kid with too much money).

I think the most recent push to go solar was precipitated by load-shedding. Under those circumstances having batteries to tide one over is essential. Load-shedding has slipped from our collective consciousness and few of us have revisited the sums. If I was doing an urban install  for a household I would have  solar geyser and a GTI capable of delivering  about 1kW with the option to add batteries and hold off on the batteries until who ever is in charge at Eskom turns the lights off.  The next day I would be in search of batteries.

However in rural municipality areas our outages are frequent and with poor quality ie dips etc there was no doubt that I wanted batteries and the ability to continue as if nothing has happened. Once you realise you need batteries then you have no choice but to get the biggest bank you can stretch to to ensure longevity. With whole house on an inverter risks to electronic equipment reduce drastically. That value is difficult to quantify.

Sent from my SM-G920F using Tapatalk

54 minutes ago, Gerald_db said:

... outages are frequent and with poor quality ie dips etc. ...

Very good point.

BUT, Axpert and Victron, most if not all solar inverters, don't handle that if grid is used. They pass most of the grid problems straight through. To handle all that can happen, is quite a mean feat and quite costly. Just check on the MLT range that they do cater for the above and things like brown-outs etc. There is quite a list. Will look for it again.

That is why I have a online UPS after my Victron inverter, protecting all my devices when on Eskom, as it is the most protected way, no direct connection from Eskom to the devices powered.

In other news. To properly sort the problems you get from grids, there are voltage rectifiers and UPS'es, cheaper than solar systems, for these problems with the grids, and it is going to get more problematic, has been around much longer than solar systems.

As per picture sync'd to the grid that is at 241v while house sits on a stable 230v come what may. Currently in self consumption mode for overnight.

Sent from my SM-G920F using Tapatalk

So after reading all your advice I have chosen to do the following.  Your advice will be much appreciated. I am going to ad another string of batteries to my system with two panels and a additional solar charger. So I will put a changeover switch between my 2 batteries strings to the axpert. When I am not on the farm the satalite Internet and phone will run off the silver calciums through the system as I currently have it. When I get there for the weekend I will switch over to the new string batteries that is connected to a additional victron so that I get the solar from the additional panel as well as the solar charge from the panels connected to the axpert. This way in theory the new batteries should last me much longer as they will only bever used a couple of times per month but they will by continously charged through the victron.  So far I am leaning towards those Trojans previously mentioned. Will my plan work. Any further advice will be much appreciated as this is going to be a additional R20k spend. Also keep in mind that when I am not there the house only needs 900wh per day and when I come in need 5500wh. Hence the two separate battery systems that I don't discharge my new batteries every day but rather use the silver calciums that I already own.

6 hours ago, Leondk said:

So after reading all your advice I have chosen to do the following.  Your advice will be much appreciated. I am going to ad another string of batteries to my system with two panels and a additional solar charger. 

Hi It is all about cost. 

I am not sure you need to replace your batteries at this stage (bird in the hand etc). What is your battery status in the morning? Looking at this post and your PV array post you definitely need more panels 500W plus some shading is not going to deliver the 5.5kWh you need when you there. Your current panels cover the 900wh base load with very little extra capacity for cloudy weather and your daily reduction due to shading.   

Depending on which SCC you buy one could for the same money get a second Axpert (the plus version) and then run two systems. Your current system with the AgCa batteries would then handle daily the daily 900Wh load.  Your new system would handle the extra load when you are there. Axperts and Trojans as a combination are problematic. There are Axperts that can handle the Trojan's equalisation charge but they do not appear to be being imported into this country.

16 hours ago, Chris Hobson said:

Axperts and Trojans as a combination are problematic.

I think Trojans T105's are very good golf cart batteries, not so sure about renewable energy applications. They like to be bashed around to mix the electrolyte around inside the battery. 18 Holes around a golf course gives them a good shakeup every day. I don't think at golf clubs they ever do an equalization charge on those batteries, as it should not be necessary. With stationary applications, like renewable energy, you have to give them an equalization charge, due to stratification and sulfation - boil everything inside the battery at a very high voltage (64.8V) for about 12 hours to mix the electrolyte inside the battery and at the same time, shed the outer layer of the battery plates like a snake sheds its skin. Every time you do an equalization charge, the plates becomes thinner and thinner. With all the bubbling and boiling at those charging voltages, you lose electrolyte. That is why those batteries have removable caps, so you can keep filling them up.

That stuff drops down to the bottom of the battery and starts to form little heaps. If the heap becomes high enough, you start to short out the plates inside the battery. I am not sure if any inverter is designed to even bulk charge those batteries at 59.3 V on a daily basis, let alone give them a equalization charge at 64.8 V every once in a while. You probably need to buy a decent golf cart charger to do that. That is just my opinion.

Since I bought T105 batteries without checking all the voltage requirements first, I am stuck with them and will just have to see how well they last with the charging voltages the Axpert 5Kva MKS inverters are capable to deliver. I cannot find a 48V charger capable of 64.8V equalisation voltage on the Internet so the only option is to find a circuit for one and build it. Just feels like way too much trouble. 

One would expect Voltronics to publish a prominent warning about the possibility that some if not all flooded LA batteries are not compatible with their Axperts.

Why don't you have look at the specs of the Microcare or another MPPT controller.  The MC can do up to 16V equalise (64V) and so it maybe easier to just run the Axpert and then disconnect it for a day every month or so to do an equalise on the bank.  You would then just run the MC on a daily basis instead of the Axpert MPPT and do the manual equalise when needed.  

I have the MC and would do this if I had trojans.

Regards

Mark

@ebrsa, those are good batteries, don't get me wrong, but unfortunately if you constantly undercharge them, you will lose capacity with time and they might not last as long, especially if you cannot equalize them either. When they talk about charging capacity, they refer to voltage, not amps. 

With daily discharge rates of 10-20%, you will not notice the loss in capacity. The day the power goes down and you are under the impression you have about 12 hours standby power, you might get a surprise that after 6-8 hours the system shuts down due to low voltage. It gets to a point where the voltage just takes a nose dive due to capacity loss.  

14 minutes ago, Mark said:

You would then just run the MC on a daily basis instead of the Axpert MPPT and do the manual equalise when needed.  

@ebrsa Yes, If you can find another MPPT controller that can charge at those voltages, I would just use that and disconnect the Axpert MPPT altogether. I am not sure how these thing work, but I am sure you can do that. 

I've done that and it works great. Currently running the axpert mppt and the microcare with 2 different strings.

It works with no hassle.

 

Edit:  After a bit of research I see the Victron MPPT's equalise at 64.8V so one spec'd to your bank may be the answer.

Just checked my batteries with the load test as explained. There is some serious problemside with a nearly 2 volt spread between the best and worst. I am thinking of buying 2 OPS260 batteries. Any opinion for best cost effective batteries?

I think the issues of undercharging Trojans needs clarification. As Plonky alluded to this in his explanation charging a Trojan battery bank with an Axpert is like driving on the freeway in one of those old 1600cc Isuzu bakkies (no I am not trying to have a fight - TTT) which can only do 90km/h whereas the speed limit is 120km/h. You will still reach your destination but not as quickly as everyone else. Even at float voltage one is still converting lead sulphate into lead and lead oxide and charging the battery.The chemical process of charging continues.  Undercharging is a function of current rather than voltage. The voltage however is critical to getting enough Amps into the battery. So what is the problem then? Well you do not have unlimited hours in a solar setup to charge using your Axpert, and that lower bulk charge may be decisive. 

If I owned Trojans I would watch my SOC critically and also watch how much current is still going into the battery during float. If I found that during float my batteries were accepting charge <1% of their Amp hour rating I would consider them full. If however this never happened I would probably charge from grid early evening until I was sure they were full. Having paid good money for my Trojans I would ensure they were being properly charged (amps being the critical factor rather than volts) even if that meant using dreaded Eskom units.

12 hours ago, Chris Hobson said:

You will still reach your destination but not as quickly as everyone else.

I agree, you can get any battery (lead acid, gel and lithium) to 100% SOC charging it at 2 amps, it might just take you 24 hours to get there. I just think batteries don't like that. When you start using that battery, it is like it has a leak in the fuel tank, the gauge drops very fast. My Samsung phone charger is in the kitchen. When I am lazy, I just plug it into my PC's USB port to charge it. At 100 mA, it takes about 2-3 hours to get to 100%. Once I take it off charge, the battery lasts 24-30 hours. If I then plug my phone into the Samsung charger at 2 amps, it is fully charged within 30-45 minutes and when I take it off charge, it lasts about 3 days.

 

12 hours ago, Chris Hobson said:

I would ensure they were being properly charged (amps being the critical factor rather than volts) even if that meant using dreaded Eskom units.

The amps required is not a problem, with the Axpert you can charge on solar (80 amps) or from Eskom (60 amps). The voltage requirement is the problem. Trojan has been in the battery business for many years. If the voltage was not that much of a problem, they would have lowered it a long time ago in line with other battery manufacturers specifications, making life a lot easier for all its customers. But for some reason, they insist on the 59.28 Volts bulk charge. It must have something to do with their technology they use inside the battery that requires high voltage for some or other process to take place. I am sure it is not a willy nilly figure they grabbed out of the air. 

If you cannot achieve the required bulk charge voltage requirements by 0.5 volt, it might not be a train smash. But if you miss it by 1.0-1.5 volts, you change the charging characteristics of the battery completely. You sit with a completely different chart than that of the manufacturer. If you put a warranty claim for batteries with unacceptable capacity and they ask you for SOC, volts and amps charts for the last 6 months to make sure you did not abuse the batteries, I think you might have a problem. 

As others have said, the Trojans are golf cart batteries. Charging them faster is important in that application, and a higher absorb voltage helps to get amp-hours in there faster. As Chris said, you can still get it done with a lower voltage, it just takes longer. In my experience, all you have to do is take a "grid day", a day when you leave all loads on grid and dump all solar into the batteries. Push it until your charge current drops to 0.5% of the capacity. If you also have battery balancers, then you can be assured that they are chock-a-block. They are sitting at absorb voltage and passing little current, ie they are full (or dead).

I usually use cloudy days for such full charges. It has enough PV to do the job and you avoid the push-pull in-and-out of the battery thing each time a cloud passes.

Also serves as a good time to sync the BMV.

Sent from my GT-I9195 using Tapatalk

The original Trojan batteries were made to service the golf cart and forklift market. The average forklift operates for 6-8 hours a day and the battery bank is sized to provide for this usage to a daily DOD of 80% and a battery lifespan of 1200 cycles (48 weeks/yr - roughly 5 years for a 5 day work week). The battery bank sizes range roughly from 300Ah to 600Ah depending of the size of the machine and its workload. Charging takes roughly 12 hours and the cycle is repeated. The weekend is a huge advantage that the forklift industry has over the solar industry. On the weekend that battery bank floats for 48 hours. A friend in the industry says if there is ever a warranty claim he looks for flat spots on the tyres. If there are flat spots it is evidence the forklift batteries having been completely discharged on the warehouse floor and the forklift being dragged back to the charging station (one would never guess that this happens in sunny South Africa with our highly skilled operators).

So in the solar industry we treat our batteries a whole lot better with DOD of 30-50% but with only 6-8 hours of charge time we probably could not have a greater DOD without a charging deficit.

There is nothing happening chemically at 59.2V that is not happening 54V.  Actually at higher voltages there is increased plate corrosion and gassing with concomitant need to top up electrolyte. A lack of gassing  could lead to stratification especially with tall cells. The magic voltage to be above is 49.6V as this is the voltage that sulphation starts to occur. The reason for adherence to manufacturer's charging voltage is to efficiently charge the batteries. When batteries are discharged there is lots of lead sulphate to act upon. As the SOC rises so there is less and less PbSO₄ at each plate to act upon the process slows down and the battery charging slows. Having a high bulk voltage prolongs the period when there is maximum current flow into the battery, an advantage when one has a limited charging time.

Trojans are different to other batteries with their thicker plates to better withstand both high voltage corrosion and the warping associated with heating during charge. Trojans might even last a bit longer with a lower charging rate due to the reduced plate corrosion rate. I have no empirical evidence for this but it stands to reason that if a charging voltage of 2.45V per cell leads to increased plate corrosion a lower charging voltage should lead to less plate corrosion.  

My not liking mixing Trojans with Axperts has more to do with the Axpert than the Trojans. One is charging at  maximum and there in not much headroom on the DC caps which is compounded by Axpert's tendency to overshoot. The inability of the Axpert to deliver an equalisation charge can be solved using one of @Chris-R's HA02s.

Thanks for sharing all your knowledge everyone who posted during the past few days, much appreciated.

My Axperts are set to charge at 58.4V, the maximum and float at 54V as per Trojan's specifications. That is only 0.88V or 1.5% below that T105 SOC is almost never allowed below 80% at which point I just manually switch to grid and set the Axperts to 20A grid charge current each. Yesterday I charged to grid to 100% SOC and then watched the charge current drop to between 0 - 0.5A with the voltage sitting at 54.0 - 54.4V. I have two strings of 48V in parallel. Each string consists of two blocks of 4 x T105s with a HA02 connected. Then another HA02 connects battery 4 of the first block to battery 1 of the second block. Both strings are of course connected the same way. After the SOC reached 100% I measured the voltages of all batteries and maximum difference on one battery was 0.1V. So I guess I am going to give up on trying to meet Trojan's specs and see what happens. @Chris Hobson it will be difficult to contradict your views and they make me feel a lot more reassured. @Don I still have to try out the Victron program but am having a busy time until Thursday. It is budget time for municipalities and I have to get a lot of stuff together for the annual difference of opinion. At least out here they still listen to the taxpayers to some extent.

I did a bit more research on LiFePo4 batteries, and it seems to me that comparisons with Lead Acid is apples and onions in many cases.

For example the cycle life. The more conservative LiFePo4 sellers spec their batteries as 2000 cycles to 80% DoD. Now the average lead-acid user is bound to do the old multiplication sum, say 5*0.8*2000 = 8000kwh, and then divide the 80k that battery costs into 8000 and conclude that at R10/kwh this is simply not worth it.

But then I found that the battery makers have what is called a warranty value. Much like a car might have a 100 000km/5 year warranty. Everyone knows the car will drive many more years after that (with the occasional issue). Similarly, what the conservative LiFePo4 makers mean with 2000 cycles, is this is the guaranteed cycle life before the battery has irreversibly lost 20% of its capacity. The battery will continue to work after that, and if treated well will give 5000 to 8000 cycles at gradually reducing capacity.

Also, unlike lead-acid there isn't a dramatic increase in aging as they get older.

When treated well, I would expect a LiFePo4 to do 2000 cycles to 80%, another 2000 to 65%, and another 2000 to 50%, or applying the distributive law, 2000*(0.8+0.6+0.5) = 3800 times the initial installed capacity. So a 5kwh bank will live for 19Mwh (and it won't be dead yet), and that 85k divided into that number brings you to around R4.50/kwh. Still not great, but better than the initial estimate.

It also seems to me that some of the trouble is related to there simply not being enough data available yet for these batteries. They haven't been around for ten years, so we don't know what to expect. Some manufacturers talk of 5000 cycles to 50%, which is once again the point at which the battery has lost 20% capacity, in other words, it could take 15000 cycles before it passes the 50% level. That brings you to R2.26, which is top end Cape Town domestic tariff. Still not quite there, as there is still the opportunity cost to take into account.

I did this math assuming 85k for the battery, which is what you'd pay for a high-end European made battery (Victron, LG-Chem, or Pylonbattery), but the local guys sell them cheaper, below 10k per kwh. If the numbers pan out -- those numbers we don't know yet -- LiFePo4 might work out as cheap as R1.50/kwh. The trouble is, that's still only good for R58 a month (2.5kwh a day at a saving of 78 cents each) whereas the opportunity cost is R250 (at 6%).

The break-even cost at present rates seems to be 5k/kwh. In other words, this starts to make sense if the battery prices halve.