How much Solar v Batteries? What do you have?

[rusty69]

4 minutes ago, peterboat said:

The americans have been using them for years in cars, so plenty of experience for them, very long lived, and they are what I am going to fit in my electric boat.

I wouldnt bother with Li-po as sodium batteries are just around the corner which because of cost, I think deal a death blow to Lithium technology

Wil be interesting to see how you get on with them. 

[Dr Bob]

1 hour ago, cuthound said:

With due respect, the bit I have highlighted shows that you don't fully understand battery charging. 

Resting voltage will show state of charge  (as will electrolyte specific gravity) but SoC is not related to capacity. A battery will show 100% charged voltage whethere it has 100% or 1% of its original capacity. 

Capacity can be measured using an amp-hour counting meter, but every organisation that I have worked for where power was critical to operation, tested capacity every two years by discharging the batteries to a test load and measuring current and time. These are the only ways to assess actual battery capacity. Also when recharging a battery you will typically need to put in 120-130% more amp-hours than you took out.

 

Er...Yes, I do understand.

The bit I have highlighted says that the point that you dont think I understand, I do understand.

Of course SoC is not related to voltage, but you dont have to be a rocket scientist to  work out how to use voltage at rest to calculate capacity. Wotever obviously understood it. I was trying not to write the ultimate detail so as to bore the guys who arent interested.

I measured the voltage at rest when full - ie 12.8 ish, discharged during the day (but not on a known load). Measured voltage at rest again and noted the amps out. You can then do what you want with the numbers - ie work out difference in state of charge between the two values and thus estimate capacity. Would you like me to show you the calculations one needs to do? It seems that you lot want all the gorey details. If your maths isnt up to it, then I can happily oblige.

I did this maybe every couple of months but mostly just noted the Ahr used to go from say 12.8 to 12.5 and if the number was the same and not trending down, then I knew the batteries were ok. I cant remember the exact numbers I used.

There are two problems with the above approach which I am sure Tony will charge in with.

1) Voltage at rest needs a couple of hours to stabilise the voltage. Not always easy. On our yacht it was more easy to do this with no inverter. With experience you could see the difference between a light load and at rest. That wasnt rocket science either. In a 'single' system, where you do the same things all the time and the number of inputs is low, it is not difficult to understand the difference between light loads and at rest.

2) Normally the accepted way is to put a known load on the batteries to calculate the Ahrs taken out. Great if you need an exact number. I was looking for a trend of decreasing Ahrs out (not in - not needed) with varying voltage at rest so an individual accurate reading was not necessary. Over a period of a year if I saw the 12.8-12.5 range use 80Ahr ( a guess as I cant remember) +/- 10Ahr, then I was happy as long as the trend wasnt downwards. Over 5 years it didnt go down much - again cant remember. I was not after an exact number. For battery charging you dont have to be exact. +/- 10% is good enough. Watch the trends and you dont have to worry too much over individual readings as long as they are near the trend line.

Whilst part of my career was spent on materials, ie plastics, paints, sealants etc I spent 10 years managing computer modeling teams, modelling a range of refinery/chemical plant applications. The models here were to predict operation of processes - viz monitor inputs and outputs and hence predict performance. Modelling a battery is no different. I made a comment in a post above about Smartguage (which I didnt elaborate on - yet it got some very negative vibes- so here is more detail). It would not be difficult at all to write a model based only on voltage that would predict a part of the charging process - After 5 years of watching batteries and charging/discharging, I could have written something similar for my set up given my knowledge of computer modelling, going from techniques such as  the simple regression models up to more challenging methods of Topology or sparse data/others.

I have read all the comments on here about Smartguages which is why I have not done any modelling myself and rely far more heavily on the Amps in/Amps out but following voltage does give you a lot of information if you use it in comparison with the other inputs. What would be interesting is to take a model based on voltage only (something similar to Smartguage) and add in the extra inputs of Ahrs change, plus inputs from the alternator (and or controller) plus solar, type/size of battery. Maybe the Ahr counters do some of this but I dont think they go far enough as they have their own model built in. Each system is diferent so you have different inputs/outputs. It would be possible to write a model for your own system and then this may be transferrable to other systems where a knowledgeable user could input his own data. Has anyone here done this?

Apologies for such a long post but I am getting tired of a number of you trying to tell me I am thick.

[Dr Bob]

2 hours ago, cuthound said:

Capacity can be measured using an amp-hour counting meter, but every organisation that I have worked for where power was critical to operation, tested capacity every two years by discharging the batteries to a test load and measuring current and time. These are the only ways to assess actual battery capacity. Also when recharging a battery you will typically need to put in 120-130% more amp-hours than you took out.

 

Fully agree but I didnt consider my application to be critical to operation. Put it into perpsective. Sailing across Biscay in a Force 6 was critical to life. A quick and dirty approach via trending can give you a reasonable assessment. You also dont need to look at amps back in. Just measure Ahr out. Would you like me to fully define how you do it by trending? No I thought not.

By the way, "Just measure Ahr out" should not be taken literally. I should have said, Just measure Ahr out and multiply by 1.25. Oh sorry, I should have said,  Just measure Ahr out and multiply by 1.25 to get an estimate of the Ahr you would need to put back in. I dont want to be criticised for writing 'shorthand'.

 

Edited July 19, 2017 by Dr Bob
more explanation for those who need it.

[Robbo]

56 minutes ago, Mike the Boilerman said:

 

I'd strongly contradict this advice for the OP as a newbie.

LiFePo4 batteries need very specific charging and management and the equipment to do this simply isn't widely available. Dedicated chargers are madatory, yet I've seen websites selling these batteries yet barely touching on LiFePo4 charging systems and equipment.

BEWARE. Charging them cannot be done with an ordinary boat engine alternator or Halfords battery charger. If fact it would be very dangerous so to do. 

I would strongly disagree.   These type of batteries are more appropriate for boaters that are off-grid and heavy users in winter when solar is minimal.

Although LifePo4 does need specific charging requirements, it's not wildly different from AGM's.  A BMS is highly recommended for all installs of LifePo4 batteries, this protects that batteries from over/under voltage and temperature.  Victron have a separate BMS, makes like Relion have them built in for a direct replacement for standard batteries.

Dedicated chargers are NOT mandatory, and you're more likely to have a charger that's appropriate anyhow.   The requirements are between 14.4v and 14.6v for charging which the majority of alternators can and do provide.   You gain best if you custom your battery charger to bulk charge at 14.6v and to disable the absorption charge as it's not required.

 

The only thing I would suggest as newbie, is to research if LifePo4 are for you, but if your off-grid in winter then they are prob the best battery type. 

Edited July 19, 2017 by Robbo

[WotEver]

43 minutes ago, peterboat said:

Look on ebay GTP batteries arnt much more expensive than Trojans

They're about 50% more expensive which may or may not matter. The maintenance-free aspect may be a big attraction. 

[WotEver]

31 minutes ago, Dr Bob said:

What would be interesting is to take a model based on voltage only (something similar to Smartguage) and add in the extra inputs of Ahrs change, plus inputs from the alternator (and or controller) plus solar, type/size of battery.

In other words, Datacell. 

http://www.merlinpowersystems.com/products/products-commercial/products-commercial-battery-monitoring/

[WotEver]

36 minutes ago, Dr Bob said:

Has anyone here done this?

See above

30 minutes ago, Dr Bob said:

I dont want to be criticised for writing 'shorthand'.

But you must consider the 'newbie' who doesn't appreciate the unwritten parts of a post and just takes any comment at face value. For instance, he might read your opening post where you stated you can use an AtoB to take batteries from 80% to 100% in less than an hour and believe it. 

[Tony Brooks]

1 hour ago, Dr Bob said:

Er...Yes, I do understand.

The bit I have highlighted says that the point that you dont think I understand, I do understand.

Of course SoC is not related to voltage, but you dont have to be a rocket scientist to  work out how to use voltage at rest to calculate capacity. Wotever obviously understood it. I was trying not to write the ultimate detail so as to bore the guys who arent interested.

I measured the voltage at rest when full - ie 12.8 ish, discharged during the day (but not on a known load). Measured voltage at rest again and noted the amps out. You can then do what you want with the numbers - ie work out difference in state of charge between the two values and thus estimate capacity. Would you like me to show you the calculations one needs to do? It seems that you lot want all the gorey details. If your maths isnt up to it, then I can happily oblige.

I did this maybe every couple of months but mostly just noted the Ahr used to go from say 12.8 to 12.5 and if the number was the same and not trending down, then I knew the batteries were ok. I cant remember the exact numbers I used.

There are two problems with the above approach which I am sure Tony will charge in with.

1) Voltage at rest needs a couple of hours to stabilise the voltage. Not always easy. On our yacht it was more easy to do this with no inverter. With experience you could see the difference between a light load and at rest. That wasnt rocket science either. In a 'single' system, where you do the same things all the time and the number of inputs is low, it is not difficult to understand the difference between light loads and at rest.

2) Normally the accepted way is to put a known load on the batteries to calculate the Ahrs taken out. Great if you need an exact number. I was looking for a trend of decreasing Ahrs out (not in - not needed) with varying voltage at rest so an individual accurate reading was not necessary. Over a period of a year if I saw the 12.8-12.5 range use 80Ahr ( a guess as I cant remember) +/- 10Ahr, then I was happy as long as the trend wasnt downwards. Over 5 years it didnt go down much - again cant remember. I was not after an exact number. For battery charging you dont have to be exact. +/- 10% is good enough. Watch the trends and you dont have to worry too much over individual readings as long as they are near the trend line.

Whilst part of my career was spent on materials, ie plastics, paints, sealants etc I spent 10 years managing computer modeling teams, modelling a range of refinery/chemical plant applications. The models here were to predict operation of processes - viz monitor inputs and outputs and hence predict performance. Modelling a battery is no different. I made a comment in a post above about Smartguage (which I didnt elaborate on - yet it got some very negative vibes- so here is more detail). It would not be difficult at all to write a model based only on voltage that would predict a part of the charging process - After 5 years of watching batteries and charging/discharging, I could have written something similar for my set up given my knowledge of computer modelling, going from techniques such as  the simple regression models up to more challenging methods of Topology or sparse data/others.

I have read all the comments on here about Smartguages which is why I have not done any modelling myself and rely far more heavily on the Amps in/Amps out but following voltage does give you a lot of information if you use it in comparison with the other inputs. What would be interesting is to take a model based on voltage only (something similar to Smartguage) and add in the extra inputs of Ahrs change, plus inputs from the alternator (and or controller) plus solar, type/size of battery. Maybe the Ahr counters do some of this but I dont think they go far enough as they have their own model built in. Each system is diferent so you have different inputs/outputs. It would be possible to write a model for your own system and then this may be transferrable to other systems where a knowledgeable user could input his own data. Has anyone here done this?

Apologies for such a long post but I am getting tired of a number of you trying to tell me I am thick.

As this thread progresses you dribble out more and more information that suggests that you know a lot more than the earlier post imply. As an example you now explain how you assess the capacity and it seem sound to me and also indicates that you know what readings on amp hour counters to trust and which to not. However along the way statements have been made that at the best are unsound so have been challenged. That is all.

For my part I felt it was necessary to say that you did not fully understand alternators, batteries, and charging because  certain of your statements were wrong or misleading. In particular the claim that an A to B can get the batteries from 80% charged to 100% charged in two hours and that the A to B reduced the charging voltage. I am happy to concede its method of operation causes the voltage reduction but its the alternator that actually reduces it. When the type of battery monitor you were using was came out the most likely and most common explanation for your beliefs (in the absence of any contra-indicators) was that you were falling into the trap such meters set for the majority of their users.

I would add that you now show that you have some degree of understanding about how battery and charging parameters vary from installation to installation and over time. All this is to the good and will show others less well informed why there is only general guidance and no set in stone rules.

 

[cuthound]

I have lots of experience with commercial battery monitors on critical applications. 

Those which show a "time left on batteries" facility generally compare falling voltage with previous stored discharge curves. All work well with batteries in excellent condition, but I have yet to find one that copes with either the coup de fout effect (where a lead acid battery that has been on float for a prolonged time suffers a temporary collapse in voltage when first put on load) or a failure towards the end of a discharge resulting from plates dropping off the group bar due to corrosion of the joint and the remaining material not being sufficient to carry the current.

For these reasons they are  viewed with considerable suspicion in the critical power industry.

[cuthound]

7 minutes ago, Tony Brooks said:

As this thread progresses you dribble out more and more information that suggests that you know a lot more than the earlier post imply. As an example you now explain how you assess the capacity and it seem sound to me and also indicates that you know what readings on amp hour counters to trust and which to not. However along the way statements have been made that at the best are unsound so have been challenged. That is all.

For my part I felt it was necessary to say that you did not fully understand alternators, batteries, and charging because  certain of your statements were wrong or misleading. In particular the claim that an A to B can get the batteries from 80% charged to 100% charged in two hours and that the A to B reduced the charging voltage. I am happy to concede its method of operation causes the voltage reduction but its the alternator that actually reduces it. When the type of battery monitor you were using was came out the most likely and most common explanation for your beliefs (in the absence of any contra-indicators) was that you were falling into the trap such meters set for the majority of their users.

I would add that you now show that you have some degree of understanding about how battery and charging parameters vary from installation to installation and over time. All this is to the good and will show others less well informed why there is only general guidance and no set in stone rules.

 

Well said Tony.

As you say, Dr Bob has, over the course of several pages of this thread, demonstrated a good understanding of batteries and their charging systems.

Unfortunately not everyone has the patience to read through long posts and those looking for a quick solution need to get sound and complete information. Without it they will continue to damage their batteries due to lack of understanding of all of the complex issues that can have an effect on achieving long battery life.

[Sea Dog]

Meanwhile, the OP (Caitlin 'cpickle' - remember her?) who confessed to limited knowledge and requested things be kept simple, is likely in the dark - both metaphorically and physically!

[rusty69]

Just now, Sea Dog said:

Meanwhile, the OP (Caitlin 'cpickle' - remember her?) who confessed to limited knowledge and requested things be kept simple, is likely in the dark - both metaphorically and physically!

I expect she will be come winter with those lighting power consumption figures. 

[WotEver]

3 minutes ago, Sea Dog said:

Meanwhile, the OP (Caitlin 'cpickle' - remember her?) who confessed to limited knowledge and requested things be kept simple, is likely in the dark - both metaphorically and physically!

I think posts 131-139 might have helped her

[Sea Dog]

4 minutes ago, WotEver said:

I think posts 131-139 might have helped her

I'd say her 136 suggests not...

[WotEver]

1 minute ago, Sea Dog said:

I'd say her 136 suggests not...

But that was answered in 137... we could go on like this for some time

[Sea Dog]

2 minutes ago, WotEver said:

But that was answered in 137... we could go on like this for some time

Aye, and I'll bet she's still in the dark!

[WotEver]

13 minutes ago, Sea Dog said:

Aye, and I'll bet she's still in the dark!

Not till October...

[MtB]

4 hours ago, peterboat said:

Look on ebay GTP batteries arnt much more expensive than Trojans and are tougher, I know which I would spend my money on.

Plus on the controller side, buy just one I had two operating together and it didnt go well [they were identical]

 

Looking at the ebay listings of these I see the seller suffers from the same problem as the sellers of those LiFePo4 batteries. Long on hyperbole and short on detail. 

In particular the only mention of how to charge them is this statement:

"Gel batteries require either close attention to charging settings or their own specialised charger."

And that's all you get on the subject of how to charge them, chaps. And they expect to sell batteries after saying that!

[Dr Bob]

3 hours ago, Tony Brooks said:

I would add that you now show that you have some degree of understanding about how battery and charging parameters vary from installation to installation and over time. All this is to the good and will show others less well informed why there is only general guidance and no set in stone rules.

 

Thank you, Tony.

[WotEver]

13 minutes ago, Mike the Boilerman said:

 

Looking at the ebay listings of these I see the seller suffers from the same problem as the sellers of those LiFePo4 batteries. Long on hyperbole and short on detail. 

In particular the only mention of how to charge them is this statement:

"Gel batteries require either close attention to charging settings or their own specialised charger."

And that's all you get on the subject of how to charge them, chaps. And they expect to sell batteries after saying that!

Fortunately Alpha Batteries are somewhat more helpful:

http://www.alpha-batteries.co.uk/12v-110ah-leoch-powabloc-tubular-gel-battery/

http://www.alpha-batteries.co.uk/media/catalog/product/g/t/gtp12-110_final.pdf

 

[Dr Bob]

4 hours ago, WotEver said:

In other words, Datacell. 

http://www.merlinpowersystems.com/products/products-commercial/products-commercial-battery-monitoring/

Well maybe, but not sure with the limited information on the link.

The problem with models (on which the Datacell will operate) is that it is not easy to have a global "fit all' model. This one seems to 'learn' over time which may make it better but is it good enough?

Voltage under load is a function of the voltage at rest and the load. Load is function of the Ahr in/out. SoC is related to Voltage under load, Capacity is related  to the voltage change per Ahr out, etc. Each installed system will have its own relationships between these and other variables. There is no reason why you cannot predict SoC and Capacity directly from monitoring voltage and load and/or the rate of change of V and Ahr. For a simple system with say 10 possible loads (lights, pumps etc) and 3 possible chargers (solar, alternator & shore power) it should be possible to come up with a 'pattern matching' type model that can predict a number of performance critera from a set of variable inputs and a set of fixed values

The problem comes when you take this model to the next boat when a lot of the fixed values (viz number and size of batteries, wire dimensions etc etc) are changed. If you are using the same model, you really need to create a new training set and determine the new relationships to get the 'perfect' predictions. All the models I ever worked with (and to be honest battery charging seems to be reasonably challenging) needed to be tuned taking them from one place to another to get the best performance.

Regression models really only work on data that is within their training sets and predictions suffer when variables and fixed inputs are outside that range - which is quite likely here. I doubt Merlin are using anything more sophisticated or they would be using it in their advertising blurb. The data logging function however looks very interesting and a good model can only be written with a good data set.

Are there any modellers in this forum that have looked this?

 

 

[Tony Brooks]

If you look back you will find the Data Cell referred to as son of Smartguage. Whilst he was still posting here Chris Gibson was developing the product and I got the impression that it used the Smartguage maps to inform an amp hour counter when to resynchronise and adjust the battery capacity. I also got a feeling that it was all taking longer than expected. It seems it was designed and priced for the military, the emergency services and such like. I really can not see most boaters paying over £1000 for battery monitoring - unless there has been a massive drop in prices since I looked.

[WotEver]

9 minutes ago, Tony Brooks said:

If you look back you will find the Data Cell referred to as son of Smartguage. Whilst he was still posting here Chris Gibson was developing the product and I got the impression that it used the Smartguage maps to inform an amp hour counter when to resynchronise and adjust the battery capacity. I also got a feeling that it was all taking longer than expected. It seems it was designed and priced for the military, the emergency services and such like. I really can not see most boaters paying over £1000 for battery monitoring - unless there has been a massive drop in prices since I looked.

That's all correct Tony. It's still around £1k, and it's the only monitor approved and used by the British (and other) military because it's the only one which is accurate.  

Most boaters are way too tight to buy one

 

[Tony Brooks]

7 minutes ago, WotEver said:

That's all correct Tony. It's still around £1k, and it's the only monitor approved and used by the British (and other) military because it's the only one which is accurate.  

Most boaters are way too tight to buy one

 

That will pay for a lot of batteries. On present form I could buy three sets for less and I expect them to last 5 years each set. I will be dead before the payback time is up!!! Or at least no longer able to boat.

[Dr Bob]

1 minute ago, WotEver said:

That's all correct Tony. It's still around £1k, and it's the only monitor approved and used by the British (and other) military because it's the only one which is accurate.  

Most boaters are way too tight to buy one

 

Absolutely, far too expensive, but that's the problem when you try and commercialise something that has to be tuned for individual use.

No, I was thinking about a group of guys doing it for free, ie like CanalPlan (which I find wonderful) - collecting data sets and seeing how to predict. I only managed a modelling team, rather than did the nitty gritty of the modelling so wouldnt be capable of the model crunching but defining parameters, data collection and model validation are the time consuming things which is within the scope of many of us.

At the end of the day, I am past all of this technical innovation and happy to see my volts high at the end of the day on my NB.