16 hours ago, andrevh said:

I need to update the firmware of the multi to proceed, but yes ESS makes especially that you can adjust it not to use all the battery as with the relay setting and Hub-1 when main available.

At the moment I am running Hub-1.

 

Battery condition:

 

Battery monitor reads 88%    Battery voltage = 49.2  and Amps being withdrawn = 30.7 Amps.

 

Is this the correct voltage or should it be higher.     ( 4 x 4 x 12 V) - 260 AH  total = 1040 AH battery bank.

There is a slight variation between different LA types but given that 49.2V  is roughly 70% resting voltage and your system shows 88% and is under load I would say 49.2V is fine. I would not be too concerned until a battery voltage below 48.5V and not under a large load.

The battery amp/voltage curve and SOC  curve  does not indicate the same here.

The battery voltage seems to fall suddenly. Can it be a memory effect?  

Maybe I need to do a recycle on all batteries,ie discharging each individual to 9.5 Volt and then charge them at 40 Amp each for 3 times?

Those 3 that had the low voltage issue did go from 10 minutes to an hour usage after the procedure.

 

 

 

 

Had the same with my Trojans. ALl would be fine and then suddenly one of them would just drop from its nominal 6V to below 5V. Battery giving notice.

Looking at the SoC chart, that straight downwards curve, that tell me the load is more or less constant. To have that sudden drop just after 8PM, that suggests you have an unhappy battery somewhere. When it happens again, get in there with a multimeter and find out which one it is.

Lead acids don't have memory effect. In fact, almost no modern battery has that anymore. Nickel-Cadmiums were known to have it, and you could fix it by cycling it fully a few times .Lead Acids don't work like that. They do get sulphated and sometimes a few cyles "reconditions" them, but it's not really the same.

@plonkster a question here, for again a bank is VRLA, which I do concede has benefits, but it cannot be equalized which "could" sort the problem with the cell/s.

Question: Are VRLA worth it compared to lead acid whom one can nurture some more to prolong life if there are cells sulphating?

Sealed batteries have that problem yes: An equalisation charge either cannot be done or has to be very carefully done. According to knurlgnar24 on youtube (its been ages since he made a video) you simply do a nice strong charge anyway. Yes, you lose some of the electrolyte but if you're lucky, not so much that it kills the battery, and besides: the battery isn't in its best health anyway so in a manner of speaking you have little to lose.

The only other option is a prolonged float charge. And by prolonged I mean weeks... :-)

Flooded cells gives you another advantage of course: You can measure the SG of individual cells and see the SG recover in the poor cell -- if it recovers at all. So maybe you are right, maybe flooded remains king. Have to tell you though, the Victron AGMs I have now (second hand ex UPS, I always add that, I would never normally buy these due to cost), they just sit there doing their job, no imbalance despite the lack of a balancer, no terminals to clean, no funny smells, I suppose those are all luxuries you pay for, but it's a real pleasure.

5 hours ago, plonkster said:

Have to tell you though ...

... that my Trojans are sitting in a wooden box in a well ventilated area with no imbalance despite the lack of a balancer, regularly equilalised because the controller is programmed to do it for me automatically, the terminals and batteries are clean, no funny smells, I suppose those are all luxuries I just got because I thought of the problem and made a plan so it's a real pleasure to have all the benefits lead acid batts can give, same as AGM's.

BUT, if I had to have them in the house, now then they would have been a problem, I do concede.

It seems there are two batteries that cannot keep the voltage. Interesting that it is always the middle batteries  (4 per bank)   (in my case) that gives the problem.

I do have three banks on 1 balancer and the recovering bank on the other. Will still connect all.

The disadvantage of the balancer is that all parallel cells are being pulled down, Without it, only one cell would have gone low.

The one bad battery, when tested on its own inverter using 350-400Watt, it lasted 13 minutes and after discharging it to below 10 Volt(unfortunately it had a sudden drop from 9.8 to 7.5 and you cannot watch it being discharged all the time) ) and then a full 40Amp (3 stage) charge, it lasted for 70 minutes at +-400 watt. Now busy with another discharge cycle.

 

 

My ESS is now fully implemented. Assuming it is allowed to return energy to the grid, what would be the most cost effective; to keep batteries charged or to use 20% of the battery as well?


Regulations require a contactor; is that not the sam as a 2 pole AC switch on the AC input line?


Sent from my iPhone using Tapatalk

39 minutes ago, andrevh said:

Regulations require a contactor; is that not the sam as a 2 pole AC switch on the AC input line?

As discussed earlier here - Regulations will require an external anti-islanding device which manages your Multiplus' connection to the grid. In the case of the UFR 1001E, the device energises or opens an external contactor to disconnect your "embedded generation" from the grid.

A 2-pole AC isolator or breaker would not fulfill this purpose. 

Plonkster uploaded a photo of his UFR 1001E box a while back - http://powerforum.co.za/topic/1134-complies-with-local-nrs-097-2-1-regulations/. going off his posts, The one relay contact (K1) manages the Contactor coil to switch the connection to the grid - I'm very sure he'll confirm if this is correct.

I'm in the process of getting my kit together to reconfigure my Victron for an AC-coupled ESS setup. A little UFR 1001E, or any other certified unit, is a very expensive, but required, part of any ESS install.

They are almost impossible to get hold of locally.

Nick

 

1 hour ago, NickNou said:

The one relay contact (K1) manages the Contactor coil to switch the connection to the grid - I'm very sure he'll confirm if this is correct.

Correct. The wiring diagram is on the Victron website.

 

1 hour ago, NickNou said:

very expensive,

True. 465 Euro, so almost 7k before shipping. Victron will charge you 35 Euro to ship it, so about 7.5k delivered to your door.

I wouldn't bother trying another unit. The ENS26 (made by UfE) is only slightly less expensive, but isn't on the Cape Town list. The ABB unit is even more expensive (and also not on the list) and the only other ones I found is for the Italian grid code (you want at least the German VDE AR 4105 if you're going to take a chance).

 

 

Quite interesting - There has been ample discussion on the topic of anti-islanding these past few months. 

Fact remains - any South African ESS install will need a UFR 1001E to be compliant. No self-engineered arrangement of contactors, coils or relays can forgo this.

1 hour ago, plonkster said:

True. 465 Euro, so almost 7k before shipping. Victron will charge you 35 Euro to ship it, so about 7.5k delivered to your door

Pricing and lack of local stock holding of anti-islanding units are part of the problem, and lends itself to the large number of non-compliant Victron installations, many of which are blissfully pumping surplus PV through old mechanical meters.

For interest: There are other options - with VDE certifications - which are better suited to very large PV plants. They are marketed as mains-decoupling relays. These include the DSE P100 from Deepsea Electronics, and the Mainspro unit from ComAp. These are not really applicable to small scale embedded generation, nor do they have NRS certificates... 

 

1 hour ago, NickNou said:

Fact remains - any South African ESS install will need a UFR 1001E to be compliant. No self-engineered arrangement of contactors, coils or relays can forgo this.

True, with one exception. The new MultiGrid inverters have built-in anti-islanding that's compliant with the German and Australian grid standards. I don't know when NRS-097-2-1 is coming, but I expect it to be soon. The MultiGrid also costs less than a MultiPlus with an added Ziehl/Contactor. It only adds about 130 Euro to the overall price, compared to the 465 for the Ziehl. Only downside to the MultiGrid: Only comes as a 3kva unit. It's meant for self-consumption grid-interactive setups and it's really adequately sized for that.

Complete agreement on the self-engineered contactor setup.

Just to be clear though -- and I'm sure the Victron owners know this -- but the multiplus has adequate anti-islanding built in already. It fails NRS-097-2-1 on the basis that there is only one disconnection device. NRS requires two switches of which one must be mechanical (not electronic). So all the Victron setups installed back in the day are completely safe. As I said before: If the Brits with their general level of nannying and health-and-safety certified it, you can be sure it's safe. What's more, the Ziehl implements passive anti-islanding. Passive anti-islanding always has a non-detection zone, a set of conditions within which an islanding event will not be noticed. To avoid this situation, active anti-islanding must be implemented by the inverter and once again: The Victron's have this, which is what allows them to be compliant once you add an extra mechanical switch.

I cannot however in good conscience tell people to skip the anti-islanding, for the simple reason that someone is going to get caught and say "but someone on the internet told me" and then they are angry with me. Rather do it right and sleep easy.

The list of alternative devices I found back in the day:

ENS26NA

Tele Haase G4PF33-1

Tele Haase G4PF21-1 (Italian grid standard)

G2FW50HzYFA02 (old German standard)

ABB CM-UFD.M33

ABB CM-UFD.M31

ABB CM-UFS.1 (old German standard)

None of them are particularly cheap, and you can't walk into your average electrical supplier and ask for one of them (even for the ABB, which has quite a large presence locally).

 

54 minutes ago, plonkster said:

The new MultiGrid inverters have built-in anti-islanding that's compliant with the German and Australian grid standards. I don't know when NRS-097-2-1 is coming, but I expect it to be soon.

True - this will greatly simplify the installation. However, I wont be surprised if most installations going forward make use of regular Multiplus and Quattro inverters regardless, even if the Multigrid is available.

From a cost standpoint, Anti-islanding add-ons just don't make any sense. An uninformed consumer won't be willing to pay R7.5k for a small blue box to comply with regulations that they might not even be checked on. Its a very difficult "concept" to sell, even at 130 Euros...

Personally, I feel that if installers and wholesalers were concerned with the above, anti-islanding devices would be readily available locally, at a reduced cost. 

42 minutes ago, plonkster said:

It fails NRS-097-2-1 on the basis that there is only one disconnection device.

Quick reading of NRS-097-2-1 states that two disconnection devices are required, of which one must be electro-mechanical and the other can be solid state.

From your points above, its clear that the Multi-plus has sufficient monitoring capabilities to perform the anti-islanding. It's a pity that the Victron can't be configured (firmware, VE configure assistant) to use one of the auxiliary relays to control a contactor as a second isolation point - this would effectively also disconnect the measurement points that actually do the monitoring.

I thought I was onto something there...

53 minutes ago, plonkster said:

So all the Victron setups installed back in the day are completely safe

Surely all installations - regardless of when they were installed and approved, need to be maintained to current requirements and standards? Even on old Hub-2 configurations, the system could be configured to export surplus solar.

38 minutes ago, NickNou said:

From a cost standpoint, Anti-islanding add-ons just don't make any sense. An uninformed consumer won't be willing to pay R7.5k for a small blue box to comply with regulations that they might not even be checked on. Its a very difficult "concept" to sell, even at 130 Euros...

The cost-sensitive customer will almost always end up with an Infini instead. I believe they are around 16k for a 3kva, while a 3kva Multi is still around 25k. And then you still have to add an extra MPPT, anti-islanding, a CCGX and sometimes a grid-meter too, easily an extra 30k on some installations. But in my experience the blue equipment is just more flexible. I know I am biased because I know the internals so well, but I would say it's like comparing the Chinese Tonka-truck replica with that expensive Lego kit: Of course the one costs more than the other :-)

I think the development of the Multigrid together with the upcoming cheaper CCGX options are important developments. The Venus GX is a Beaglebone Black based replacement for the CCGX, that will cost around 4k. But you can also run the software on your own hardware (BBB or Rpi), that option just comes with no support. The support is the expensive part you see. They are also replacing the expensive Carlo Gavazzi EM24 with a cheaper one (about half the price I think). So the 8k CCGX comes down to 4k, the 4k CG comes down to maybe 2k, and the 7.5k Ziehl is now a mere 2k. That's almost 12k down from what it was...

Is it enough to convince the Infini customer? I suspect not. But then, I'm also pretty sure the blue stuff will still be around in 15 years, but I don't know about the China boxes... :-)

8 hours ago, andrevh said:

My ESS is now fully implemented. Assuming it is allowed to return energy to the grid, what would be the most cost effective; to keep batteries charged or to use 20% of the battery as well?

Some principles I live by:

1. Don't take power out of a battery unless you know you can put it back tomorrow.

2. Don't take power out of the battery if you can buy it cheaper from the grid.

This pretty much means that it is best to keep the batteries charged or cycle them very shallowly.

But I also found that if you keep the batteries at 100% all the time, you lose out on some self-consumption. So what I do at the moment is to set my Minimum Soc to 95%, Optimised, no Battery Life. So I literally use the top 5% of the battery just to do a bit of bridging. This works for me because my array is -- to put it bluntly -- grossly undersized compared to my consumption, so I can use (almost) everything I produce straight away.

In other words, because I have very little surplus, I cannot necessarily put back 20% if I take it out, and because battery cost is still a couple of cents above even the most expensive grid electricity, there is certainly very little reason to cycle my batteries any deeper than the bare minimum.

Once my array is upgraded (need to find the time, the panels are standing here just waiting), I'll have a new problem. I'll have to decide if it's best to cycle the batteries, or not. Difficult question really, what you want to do is set it as high as possible and use as much as possible immediately.

Also, the "Keep Batteries Charged" option is broken at the moment, or it is on my setup. Might have something to do with the canbus controller. So I have very little choice, I have to use one of the other modes. Optimised mode with a very high Minimum SoC works well for me.

Hi

Just some feedback on the system after I had all banks on battery balancers and some recycling of batteries. So all batteries are still the same that produced the above curves.

The current current curves seems better.