How bad is it to use Battery volts for State of Charge

[PaulF007]

I have been looking at the Graphs that Emon Produces and one cool graph that you can use is the comparison graph. Basicly it takes two feeds and compare it to each other and then plot it on a line.

So I compared State of Charge with the battery Volts from the BMV 702 and this should give you an idea as to how bad it measures.  

First is the two feeds overlayed on top of each other and as you can see although the Soc changes the voltage stays more or less on the same average

 

 

Emon plots the comparison on a straight line and there is no similarity between the two - Y axis = State of Charge and X axis = Battery volts on the same time

 

Just as a base I compare SOC with AH used and they are directly related as you can see here.

 

Would be interesting to see how the data compares with some other emon users.

 

[Mark]

Is this local emon or .org. Excellent analysis tool. Love graph 3.

Sent from my SM-G800F using Tapatalk

[PaulF007]

It is on the local Emoncms Software.

[stuvo]

I was just about to create a new topic on whether or not I really needed a BMV-700. You seem to be saying YES I do. I'm still a bit unclear on how I would use the information provided by the monitor. Will I need to manually change the charging and battery settings daily based on the BMV?

 

Equipment:
SMKSOLAR Axpert MKS 5kVA 4000w 80A Inverter / Charger
8ea.  Rocket AGM Batteries 12v 100ah. Wired as 2 parallel strings of 4 batts (48v)
12ea. Seraphim 260w panels. Wired as 4 parallel strings of 3 panels (3.12kw)

PROGRAM:
1      Output source priority               SbU
2      Maximum charging current       50A
3      AC input voltage range              APL
5      Battery type                               AGM
11    Max Utility charging current       10A
12    Set V point back to util source    49v
13    Set V point back to batt mode    FUL or ?default 54V?
16    Charger source priority              CSO
31    Solar power balance                  SbE

[Don]

Hi @stuvo. The general consensus is that you do need a BMV. If you have to buy one, get the BMV-702 model. The reason why you need a BMV is that it is the only way to accurately determine the State of Charge of your batteries. You can setup alarm levels that would alarm when certain levels of charge or discharge have been reached that is undesirable. It keeps a detailed log of deepest discharge, lowest discharge volts, cumulative kWh charge and discharge, and many more. 

As I said before, the main reason is to accurately determine your State of Charge. I am not sure what monitoring software you are making use of. As an example, if you are making use of WatchPower, your State of Charge is a calculation based on a volt reading which is highly inaccurate. 

53 minutes ago, stuvo said:

Will I need to manually change the charging and battery settings daily based on the BMV?

 

No. If you for example make use of monitoring software like ICC for Windows or Raspberry Pi, it integrates the BMV data with the Inverter data, all into one solution. You can then select settings, either time based or based on you State of Charge, when to switch to solar or to grid.  

Just as important as a BMV, are battery ballancers. The guys on the forum have had great success with the HA-02 battery balancers. If you search the forum, you will find many threads about them. If you are unsure, feel free to ask. 

[Chris Hobson]

These graphs are really cool. The one graph shows the charge/discharge cycle quite clearly.

If we start at the bottom left the green line shows the relations ship between SOC and voltage during bulk phase. It is not a perfect relationship but could be described by a linear function. Solar charging is imperfect in that there are periods of lesser charge and maybe even periods of discharge thrown in. If this charging were to take place as grid charging I would expect a much more linear relationship. 

The orange line describes absorb where there is no further increase in voltage (well there is but we dealing with an Axpert) and an increase in SOC. 

Float is the blue line and goes from about 56.2V all the way to about 51.2V  when the batteries start discharging, with a concentration of readings around 54V which the the voltage at which float has been set.

The yellow line describes the discharge part of the cycle and the only time when one could in essence use voltage to determine SOC. It's not perfect but there is clearly a linear relationship. We tend to forget that the relationship between SOC and battery voltage is for a rested battery and with a renewable energy system you just don't have a rested battery unless you disconnect your battery bank. The closest we come to this is a battery bank that has a small discharge and that we get mostly overnight once everyone has gone to bed.

I do find myself making quick SOC calculations using battery voltages when I am visiting  folk who do not have a BMV.  You do have to take into account the fact that the battery is not at resting voltage and under light discharge the resting voltage would be about 0.2 to 0.5 V higher than the connected voltage reading.

That voltage will then describe SOC according to this table

And if we look at Paul's SOC reading bottom left on voltage and looking at the table we come to a SOC of somewhere between 80 and 90%. Not perfect but nevertheless some indication of SOC in the absence of a BMV reading.

A BMV is essential to my mind. Unfortunately they are quite pricey but in terms of a insurance policy for you batteries I think money well spent. I would spend the extra money and  buy a 702 so that you can do midpoint monitoring. I am convinced that if I had not spent the money on a BMV I would be looking at replacing my batteries prematurely. The BMV alerted me to the fact that I had imbalance in my battery bank that I then remedied with the purchase of a battery balancing system. VRLA batteries do not take kindly to an equalising charge and hence my preoccupation with battery balancing. 

[Cef]

On 10/3/2017 at 4:10 AM, Chris Hobson said:

...

That voltage will then describe SOC according to this table

...

Hello Chris. Sorry for the question, but I'm confused.
I was able to operate my system, among them there is a parameter corresponding to Program No. 19 "Low DC cut-off battery voltage Setting" that I can not understand in what value to fix so that my batteries do not get discharged more than 40% (theoretical SOC 60%). My inverter allows a Range of between 40.8 to 48.0 Volts with 0.1 volt intervals. What would be the value (Volts) that you would place if you wish that your batteries do not discharge more than 40%? Thank you very much.

[Chris Hobson]

 

 

7 hours ago, Elmichi said:

I was able to operate my system, among them there is a parameter corresponding to Program No. 19 "Low DC cut-off battery voltage Setting" that I can not understand in what value to fix so that my batteries do not get discharged more than 40% (theoretical SOC 60%).

I think you mean program 29. Unfortunately it is not something I can give a straight answer to but I will provide you with the information so that you can answer your own question.

• Battery bank size together with high discharges has an influence on voltage fluctuations. A larger bank has a moderating influence on the variation caused by high discharges. A large draw causes battery voltage to dip more than would be expected due to the discharge. After you have removed the large load battery voltage recovers but obviously not to the level before the large load occurred.
• So if you had a large load regularly and you have a small bank then voltage is going to be a very poor approximation of SOC. So setting program 29 to its maximum of 48V at best would be a SOC of 50%. One has to be cautious though - You are going to be switching your inverter off. A fictitious example. You have worked very conservatively with your battery bank and the voltage just before sunrise is 49.0V which is equivalent to SOC of about 75% on a battery under light load. Your missus switches on the kettle to have a morning cuppa and the voltage drops quickly to below 48V and your inverter switches off. Your missus has a cold cup of coffee and you get hot tongue and cold shoulder. If your battery bank was bigger the voltage drop would not be so severe and you might get away with it.

My battery bank is small so my battery cutoff is set currently to 46.V. With me logging minimum voltage I am better able to see what is out of the ordinary and may raise it. I actually rely on a BMV's relay to switch off the load at 55% SOC and use program 29 as a backstop.

[Cef]

Many thanks @Chris, as always willing to help and so clear in your answers.
I confirmed that the program number is [#19]. My inverter is a MPP Solar Hybrid V of 4Kw and the program numbers must be different from those of Axpert .
For the anxiety of starting my system I have not yet connected the BMV-702, but I will do it as soon as I can.
While we all know that it is possible to calculate what I am going to ask you mathematically, and based on what you comment in the sense that you have a small battery bank (I see that it is 260 Ah), mine is even lower (225 Ah ) And because of your experience of use, I ask you what discharge rate (in Watts) is the time it takes for your BMV to send the cut signal from your fully charged battery bank in float state, for example with a Consumption of 1,000 Watts, having adjusted the BMV at 55% SOC?

[Chris Hobson]

You can work it out. I show you how. We are going to ignore losses which obviously is not the right thing to do but you wanting ballpark figures and we are going to be within 10% or so of the true value.

OK 225Ah battery bank is going to give you 48×225 =  10 800 Wh (and yes the battery is not at 48V but we doing ballpark figures). 45% of this is 4860 Wh. With a load of 1000W that is going to last you  4.86 hours. Lets say just over 4 hours to include losses. Not enough for overnight but enough to power you through an average power outage or load shedding. 1000W is serious draw for overnight. I manage to have a overnight draw of 250-500W with an average of about 350W. So if this is really your draw then first reduce the draw. Turn off the ice machine at night or whatever is drawing a large amount of current. I still think 225Ah is too small to go through the night so I would use grid power if it is available at night and save the batteries for when you need them. You might want to cycle them lightly to get some use out of them. If you are off grid then I would say two strings of batteries minimum. My battery bank is not enough to see my household through the night. If I am here on my own it's fine. Normally the gennie runs for 3-4 hours to provide power to my employees in the evening and we piggyback on that to see us through early evening and then my bank lasts through the night. The next project is to provide them with power and I will have to double my battery bank but as the saying goes one eats an elephant piece by piece.

[Cef]

1 hour ago, Chris Hobson said:

You can work it out. I show you how. We are going to ignore losses which obviously is not the right thing to do but you wanting ballpark figures and we are going to be within 10% or so of the true value...

Yes Chris, are the calculations that I had, I do not know if there will be any problem in my batteries  ... I started my system on Sunday 26. The new Trojan batteries began to load immediately (Trojan recommends "equalization charge" to New batteries but unfortunately the limitation of my current inverter does not allow me to do it, for that reason I am thinking of buying a specific MPT Charger Midnite or Victron ... but that will be a reason for a future post  ) .
I set the load current to 20 Ah in the corresponding program. Unfortunately the Voltage of my inverter does not reach the one recommended by Troyan for the "Bulk" load, but I set it in the highest possible value ...
I left for about 24 hours charging the battery bank until it reached the float voltage.
I had set the voltage of CUT OFF in program [19] to 47.0 Volts.
I have an Eastron 220 Mod Bus Power Meter and before simulating a grid power failure I noticed that the consumption during the last 5 minutes was between 900 and 1050 watts of active power.
I was surprised to note that once I made the cut of the grid Energy, there were no more than 7 (SEVEN) !!! Minutes until the alarm of imminent cut of my current inverter began to ring ...
I'm going to set the program [19] to 46.0 Volts.

Thank you for your comments !

[Chris Hobson]

33 minutes ago, Elmichi said:

 Unfortunately the Voltage of my inverter does not reach the one recommended by Troyan for the "Bulk" load, but I set it in the highest possible value ...
I left for about 24 hours charging the battery bank until it reached the float voltage

Plonkster liken this to being on the freeway where the speed limit is 120 km/h but your old jalopy can only do 90km/h you will eventually get there. Your batteries will go into absorb earlier and stay in absorb longer.

37 minutes ago, Elmichi said:

.... unfortunately the limitation of my current inverter does not allow me to do it, for that reason I am thinking of buying a specific MPT Charger Midnite or Victron ... but that will be a reason for a future post  ) 

I would recommend this. You have very good batteries and you want to charge and equalise to the specifications of the datasheet.

 

43 minutes ago, Elmichi said:

I was surprised to note that once I made the cut of the grid Energy, there were no more than 7 (SEVEN) !!! Minutes until the alarm of imminent cut of my current inverter began to ring ..

I will check this evening but the alarm I think begins to sound about 2-3 V above the cutoff so this does not exactly surprise me.

[Cef]

Chris Hobson, I thank you again.

 

2 hours ago, Chris Hobson said:

 

I would recommend this. You have very good batteries and you want to charge and equalise to the specifications of the datasheet

I wish I could do it !!!. The MPP Solar Hybrid V has no EQ mode !! I think I'm going to buy a dedicated MPPT Battery Charger in a short time ...

Now I am on a work trip away from my home, I come back next week and I will immediately program a total cut off voltage of 46.0 and try again. The truth was that I had no idea that the alarm was starting to sound 2 to 3 Volts above the REAL Cut Off. Best regards !

[___]

14 hours ago, Chris Hobson said:

Plonkster liken this to being on the freeway where the speed limit is 120 km/h but your old jalopy can only do 90km/h you will eventually get there. Your batteries will go into absorb earlier and stay in absorb longer.

LOL, I made quite an impression :-) As long as you get more or less to a voltage that is sufficiently high that it drives the chemical conversion in the battery the battery will eventually be fully charged, it will just take longer to get there. The caveat is that the voltage must at least be high enough to qualify as a "gassing" voltage, and obviously the further you are off the longer it will then take to get there. You can recharge a battery at float-voltages too, it will just take a week to get there! :-)

In fact, that's precisely what some UPSes do...

[Chris Hobson]

42 minutes ago, plonkster said:

LOL, I made quite an impression :-)

I love analogies 

That was a gem along with Don's cricket analogy and the two stroke 4 stroke diesel analogy. It is always easier to explain the unfamiliar by relating it to the familiar.