My installation started in June 2019, but the actual process started in February 2019.

The installation:

• Victron Multipuls II 48/5000
• 1xVictron MPPT 150/85
• 1x Victron Color Control GX
• 2 x US3000 Pylontech Li-Ion Batteries
• 16 x 330w Tier 1 Solar Panels
• 1 x ET112 Energy Meter
• 2 x Timers for my geysers
• 1 x "new" Non-essential DB

I chose Victron mostly of what I read here on the forum, and also because I felt the Invertor is the best "value" for money compared to the more expensive "German" makes and cheaper "Chinese" makes and it is extensible if you need it to be. There is probably other makes I could have chosen, but I felt most comfortable with Victron, if that makes any sense.

I started by getting quotes, but I did ask for the same specs, otherwise you cannot compare apples with apples. Also I am not really a DIY type person so I chose to get a installer, but finding one is not so easy. I got some references and eventual decided on cost vs experience, I chose cost 😉, which brought it's own challenges but it was my choice. They had some issues in getting the install 100% but I never felt that they would abandon me, during the process, but my patience was tested a few times.

My aim was/is to, reduce my Electricity cost a month as well as have a backup in the evenings "if/when" we get load shedding again. Yes I know it is still cheaper to use Eskom compared to solar with batteries, but I am willing to pay for that. Where I stay, we pay a sliding scale of 0-600kWh at a cheaper rate, then from 601kWh and up we pay a more expensive rate. So the aim is to try and stay below 600kWh a month. My avg consumption over a 12 month period last year was 890 kWh, with a max of 1330kWh in July. I am also allowed to feedback to the Grid where I stay as an added benefit.

Which brings me to the community here, after asking a question in the "Starting In Solar? Feel free to introduce yourself" section, after struggling with the installer for almost two weeks in getting the finer details correctly setup. @Jaco de Jongh pm me to assist, and offered to come to my house, in getting the install working 100% at not cost! He even phoned @plonkster on a questioned I had regarding battery usage. I just want to again say thank you, to BOTH of them, in assisting in getting it working and explaining it to me in a manner of how it actual works and what I "thought" it is suppose to work like.

 

Some important learning usage as a newbie:

First thing that I learned is that when your geyser is on and connected on your non-essential load DB, but the Grid is still available, ie no power failure, then it will still use the battery power to operate, if no/some PV (solar) power is available.

Let me explain: Say for instance it is 17:00 in the afternoon,  your PV generation is generating 500w but you have set your Geyser to switch also on at 17:00 to warm the water for two hours to be able to to bath/shower your children etc, and your battery is fully charged from your PV system, your usage will then go to say 3500w for your house , it will first draw 500w from the PV, then it will draw the remaining 3000w form your battery and or Grid, which could means it could draw 2000W from your battery and 1000w from your grid, which in theory can drain your battery very quickly, by 4 kWh, in two hours, then at 19:30 you have a power failure, then you have barely any power left to use. To overcome that you can setup schedule charging in ESS, on your CCGX, which will allow not to use the battery unless there is a power failure. For it to be effective you also need to set the SOC, to a lower value than your current SOC.

I hope this explanation makes sense. (Welcome to correct me if I am wrong)

What I have done is to setup schedule charging to not allow for the battery to discharge when my geyser is switch on and then also late afternoon early evening when you use the microwave and or oven for cooking etc. I fully start using the batteries at 20:45 at night, and so far I have been running my whole house until 05:00 weekdays, 06:00 weekends, in the mornings on battery power, with some reserve left on my battery of about 40%, and then it uses that for the geyser to switch on at 05:00 on weekdays and at 06:00 on weekends. This whole process has taken me about a month to figure out what works best for our household. You would say a month?, yes, because in the beginning the whole household, needs to adjust to using solar and if you have teenagers in your house... then well you have to have, patience...😃

 

Edited August 19, 20196 yr by isimobile

1 hour ago, isimobile said:

16 x 330w Tier 1 Solar Panels

 

1 hour ago, isimobile said:

1xVictron MPPT 150/85

 

What are the specs of those panels? 

I have 18 x 275W using a 250/100 MPPT. and i am very close to the limit.

Oh, wait... i just realised i spec'd the system with more panels and in the process i forgot that i actually bought less panels. Thx to this post i now remember that i have space to add more panels

Edited August 19, 20196 yr by stoic

Very Neat. Well done. I'm Blue with envy!!! Coz I only wish I could offered Blue.

To Power forum Team they are extremely helpful is every way. They are so called a few men apart!!! Very hard to find in today's time.

1 hour ago, isimobile said:

. @Jaco de Jongh pm me to assist, and offered to come to my house, in getting the install working 100% at not cost! He even phoned @plonkster on a questioned I had regarding battery usage.

He was quite lucky that I had working internet at the time! I Was in a neighbouring country and whatsapp was the only thing we could use. The contention ratio on the nearby mobile tower (that also runs the farmers' WiMAX connections) is too high and during the day it's hit and miss. But we help where we can... 🙂

@plonkster
I was thinking, to overcome the discharge of battery when using a high powered device, like a geyser & oven, how about adding a max discharge value for battery setting in ess, when you hit it, it switches to a "scheduled charging mode X" with  grid available?

So pseudo :

 If (grid available and current usage > than threshold set in ess)

     Set schedule charging mode X on

     do normal

Else

  Switchoff schedule charging mode X off

    Do normal

End

Just an idea? 🤔.

 

 

 

27 minutes ago, isimobile said:

I was thinking, ...

I'm thinking of enabling NodeRed and manage it there, as I want to use less batteries yet all the panels with high powered devices yet have 100% SOC at say 5pm for 12am-8am use.

1 hour ago, isimobile said:

max discharge value for battery

Here is my suggestion. If enough people ask for it, then maybe it will happen. But you have to ask for it using an official platform, like the Victron community site. Otherwise it's just anecdotal evidence. I hear the request every now and then, but I know literally of less than 5 people who want it...

Second, it's not easy to do. You may think that it is, but the main obstacle is this: The Multi has no DC current measurement (and it's not unique in this sense, many other inverters also don't have it). DC current is estimated from the AC values. In theory the DC power is the AC input power minus the AC output power, multiplied by an efficiency factor (0.9 is assumed), and then the result is divided by the DC voltage and that gets you the DC current. For values above 2A, it's not too far off, but especially under 1A  it's practically noise.

Now imagine someone ask for a max BATTERY discharge of 0A. In other words, he wants to use his PV, but the battery should not discharge at all. Now imagine that due to our inability to estimate the DC current, we "accidentally" discharge the battery slowly.

And this will simply be the case in any system where the PV is DC-tied and you don't have accurate DC measurements.

Now compare the number of people asking for the feature with the number of people who will complain that it doesn't work correctly.

It is far better to let people solve this using their own software platforms... UNLESS enough people ask for it.

5 hours ago, isimobile said:

2 x US3000 Pylontech Li-Ion Batteries

Just as info. The battery's is actually not connected to recommendation. The black(negative) wire from coming from the  inverter should connect to the negative pole on the bottom battery. See illustration below.

I know on two batteries the charge/discharge imbalance might be minute if you keep it as it is, but something like that will drive me crazy eventually and I would end up fixing it.   

 

1 hour ago, plonkster said:

it's not easy to do. You may think that it is

Nope, would never think that it is. 😉

I also get regalar request in programming todo something at work, because it is easy, l always say log a request, and it will be investigated. Even if the requester(usually a manager or The boss) think it is a 5 minute task, it ends up 10 hours of work... and sometimes not... 😂. Then 5 bugs later after testing, because everything is "related", you think why do people think sometimes 1 liner requests is simple todo. 🤔.

As you put it, the details seems very intricate, and you need to understand the complete design.

1 hour ago, plonkster said:

UNLESS enough people ask for it.

I will put the suggestion on the official community.

3 hours ago, plonkster said:

The Multi has no DC current measurement ...

I am one of the people, there are a couple more on Victron site.

1 hour ago, isimobile said:

will put the suggestion on the official community.

Let me know where, and I'll join.

My suggestion to solve the DC current measurement is that one has to have a BMV. If not, you HAVE to accept anything not perfect. It is one of these NOT NEGOTIABLE points in all software development.

EDIT: Some may say it is easy to say when you have a BMV, yes, I agree, it is. But WHEN I see the not perfect solution, and I know I want it right, I will spend the monies, especially seeing that a BMV on lithium batts adds a better SOC than a BMS only.

Edited August 19, 20196 yr by Guest

7 minutes ago, The Terrible Triplett said:

one has to have a BMV

Even then, you're talking about a closed control loop, where the output is fed back into the control loop in order. Worse, because there are measurement delays, you're feeding back an old measurement into the loop. Closed control loops always have a real danger of going unstable.

I had a long conversation with TTT the other day about how cars also have a closed control loop and an open loop. At lower RPMs and idle (once the engine is warmed up) the lambda sensor (oxygen sensor in the exhaust line) is used in a closed loop so the air/fuel mixture is continually adjusted for the least emissions. Closed loop, the output literally feeds the input.

But, put some lead in your shoe and it tosses the closed control out the window, goes open-loop instead and makes the mixture rich. Why? Cause it makes more power, and a rich mixture is safer than a lean one. This particular case seems so similar to me. Again, at large power values, it doesn't matter if you run a little rich (a few amps over). It's the low-power conditions that are hard to do!

Edited August 19, 20196 yr by plonkster

2 minutes ago, plonkster said:

It's the low-power conditions that are hard to do!

Maybe an idea to have a min low-power value?

2 minutes ago, The Terrible Triplett said:

Maybe an idea to have a min low-power value?

Technically, if you need your battery charge/discharge to be that accurate, then feed your PV on the AC side. Regulating on the DC bus using widely spaced measurements in a closed loop seems like a bad idea. You'd have to damp the heck out of it.

In fact, I think it might be easier to fit some kind of function, possibly a linear regression would do it, and simply make it "better" and forget about perfect. Thing is... it is complex enough as it is.

22 minutes ago, plonkster said:

Thing is... it is complex enough as it is.

Then leave it I say. I'll see if NodeRed does the job I want done better.

8 hours ago, The Terrible Triplett said:

Then leave it I say. I'll see if NodeRed does the job I want done better.

It's not only the discharge side that is difficult. The charge side has similar challenges. Some batteries really want you to hit precise charge current limits and if you don't, they disconnect. This one particular battery, which I will not mention to protect the guilty, throws alarms if you're off by even a few ampere, even if you lower the charge voltage. Now imagine there's a big solar array on there with DC-tied MPPTs and you want to exactly balance the inverter DC current with the PV current so the remainder is zero and the battery doesn't get any of it... well with PV power fluctuating slightly even on the best of days and loads doing the same... there's just no way you can do it exactly. Which again leads me to the conclusion: If you need to do it that precisely, the PV needs to be AC-tied so that there is a single point of control for battery current (the Multi). For DC-tied PV, you should use a battery that regulates the charge voltage rather than the charge current.

57 minutes ago, plonkster said:

It's not only the discharge side that is difficult.

Ok, I am now totally convinced that this idea some of us have, as cool as it might sound / be, is not really that cool an idea to implement due to complexities and other things that may fall off the cart. Anybody who is in or around development knows that nothing is as easy as it seems, ever.

58 minutes ago, The Terrible Triplett said:

totally convinced

As long as you can use the voltage, you can make it work. I've done things like that with lead acid, regulate on the voltage. If the voltage goes over 13.5V (technically a gassing voltage), then increase the discharge limit, if it drifts under 13V, then decrease it. Successively doing this makes it average out around a point where the battery is getting a good float charge. This is essentially what Keep Batteries Charged does anyway. It even works perfectly well with LiFePO4, but not with all BMSes. That is where the spanner in the works lives.

There are tricks you could try. Assume again the scenario where the user has set the maximum battery discharge to 0A. You could use a pessimistic estimate of inverter efficiency so that on average the battery is being charged slightly. That works... up until the point where the battery gets full. When the battery gets full, the solar chargers pull back, and because we're using a pessimistic estimate of efficiency, the amount we feed to the loads also pulls back by the same ratio, leading to a PV death spiral.

So another idea I've messed with is to switch between a pessimistic and an optimistic estimate based on how full the battery is, but this has problems too. Remember that SOC is always an estimate, and it drifts... and at lower current values (which is what you will get when you aim for 0A) it is spectacularly wrong. So this idea leads to terrible SOC drift in some cases.

Also, any idea that causes a slight charge/discharge cycle or a sawtooth voltage chart... leads to complaints from (some) customers 🙂

 

5 minutes ago, plonkster said:

... sawtooth voltage chart ... leads to complaints from (some) customers

Customers, they are always right ... until proven wrong!!!

Complaints like what you just mentioned coupled with the infatuation of the world today that a customer is ALWAYS king, ALWAYS right, must be obeyed ... his/her money is worth more than yours / your time ... it goes 100% against what our wife's tell us: We are not always right. 🤣

Note: I am NOT referring to any customer complaints that have solid merits as in they are caused by idiotic suppliers / sellers / thieves / customers that are lied to. 

I always will bend backwards for my clients but if they are wrong, then it is my duty to patiently explain that to them ... that is if I want to retain them.

So those clients who complain about a "saw tooth voltage chart" ... makes me want to ask: Do you want them? ( NO!!! The customer is not always right. )

20 minutes ago, The Terrible Triplett said:

Customers, they are always right ... until proven wrong!!!

The idea that the customer is always right was never intended to be some absolute rule. It means merely that in as much as it is possible to do what he asks, even if you disagree with the combination, you make him what he asks for. If the customer wants to have a bright pink paint job on his car, and you have pink paint in stock, then you spray his car pink regardless of what you think about it.

Even this principle isn't absolute. In some parts of Italy, if you ask for banana on your Pizza, the Chef will tell you that he thinks that won't go well and refuse to make it. Some French restaurants will refuse to serve the wrong kind of wine with some dishes. And so forth.

Nevertheless, the principle (where it applies) only applies in as much as it is technically possible in the first place. Beyond that point, the ENGINEER is always right!

Original here.

13 minutes ago, plonkster said:

The idea that the customer is always right was never intended to be some absolute rule.

Let me guess. You have not recently worked as a waiter or in retail? 
It is bad bro, very very bad. What gets my heckles up to a point where I actually say something is when a customers goes for a teller in like Spar / PnP etc. Man, that is just plain bullying. 

Customer Bullies go for the lowly paid staff time and time again hiding behind "The Customer Is Always Right".

Respect for those chefs and waiters who refuse service !!! More should do that.

 

13 minutes ago, plonkster said:

Nevertheless, the principle (where it applies) only applies in as much as it is technically possible in the first place. Beyond that point, the ENGINEER is always right!

Amen ... but hold on, even that limit gets pushed. Just ask my Father in Law. He, being the engineer, at times walks away from a demanding customer who in the end "gets what they paid for" ... and he gets called back and O'boy, then that bill just went up.

 

Back to the thread. Point I am making is that on the limiting of the current on a ESS system, the app cannot to "identify the birds". 👍

Edited August 20, 20196 yr by Guest

An example of how far customers push the limits. I know this handyman. He is damn good.

So he gets called to a very expensive house in a very expensive area by a very rich lady to install a bunch of very expensive things she bought for a party the next evening.

Party is over, he gets called back to remove it all and put it back into the boxes. The lady is returning the goods.

The commission that low paid retail staff member got on that sale, wiped out, having had to accept the returned goods, all in the name of Customer Care. 

Made me sick to my core.

3 minutes ago, The Terrible Triplett said:

Made me sick to my core.

Second hand building material store here in Strand. I am looking for a replacement cistern. I find one... without lid. I really want the lid, because mine isn't looking so good... but the cistern itself is cracked so that one is more important. I ask the owner of the store: Where is the lid? Oh, someone borrowed it. They'll bring it back later. They needed it for an inspection (someone's lease ended and they don't want to pay for breaking the lid).

I'm a landlord. I almost told him that too...

@plonkster, I see a similar question was already asked in the Victron community, https://community.victronenergy.com/questions/212/ess-suggestion-for-better-battery-use.html, about max discharge, Have you seen it before?

Edited August 27, 20196 yr by isimobile