Running a set of 12V Trojans down

[lw395]

Oh sure analogies are never perfect and certainly my last paragraph in particular is a considerable oversimplification.
However I think the first part gets over the important point that by discharging a battery more slowly to a given end voltage, you are not increasing its real capacity, you are increasing the depth of discharge relative to nominal capacity (i.e. you are converting a higher proportion of the active material) and the cycle life effect is commensurate. That's the point under discussion here, not recharge efficiency.
There's no such thing as a free lunch.

AIUI, there is inefficiency in discharge too, so the capacity you get out varies according to the rate at which you discharge it.
I get your point that if you take out say 50Ah at 50A, the end voltage will be lower than if you take out 50Ah at 5A. Because the 50A will be causing a significant drop from resistance etc. But if you let the battery settle for a day and then drew 5A in both cases, the battery which had be drained at 50A would generally have a lower voltage than the other.
At the other end of the scale, there is self discharge, so draining a bettery slowly can be inefficient too.
I believe some batteries, drawing a small current can increase the self discharge. I don't know if that's true of lead acid.

 

[lw395]

Let’s look at the AH capacities of a common solar battery, the Trojan L16 6 Volt battery. According to Trojan, the ratings are as follows:
Battery "C" Rating Battery Capacity Load Available Energy
C5 HOURS 303 AH 60.6 AMPS 1.82 KWH
C10 HOURS 340 AH 34.0 AMPS 2.04 KWH
C20 HOURS 370 AH 18.5 AMPS 2.22 KWH
C100 HOURS 411 AH 4.11 AMPS 2.47 KWH

From:
http://solarhomestead.com/battery-amp-hour-ratings/


I thought this was interesting too:
https://www.researchgate.net/deref/...cimages/7427/Lead_Acid_Battery_Efficiency.pdf

As a battery gets to higher charge levels, the charging efficiency plummets.
"These tests indicate that from zero SOC to 84% SOC the
average overall battery charging efficiency is 91%, and
that the incremental battery charging efficiency from 79%
to 84% is only 55%. "

So having batteries bigger than you need can be very inefficient.

 

[Mistroma]

See comments interposed in your quoted message.



I suspect you are still thinking of the battery as a device that stores electricity, akin to a capacitor.
It isn't.
It's a chemical reactor.
During discharge you consume chemicals to produce electrical energy. Producing 1Ah of electrical energy will always consume the same amounts of the chemicals - lead oxide at +ve plate converting to lead sulphate, metallic lead at -ve plate also converting to lead sulphate, and sulphuric acid becoming more dilute - regardless of reaction rate. The "efficiency" does not vary. That electrical energy is then dissipated partly in the internal resistance and partly in the external load. The voltage reduces also as the acid specific gravity reduces during the discharge so a constant current is not a constant power.

(During recharge, the process efficiency to regenerate the chemicals does vary a lot according to charging conditions and state of charge, that's an entirely different case.)

Courtesy of Wikipedia (to save me scratching my head to remember) -

EDIT - dammit all the formatting has gone, look at it here https://en.wikipedia.org/wiki/Lead–acid_battery in the section headed "discharge"

Note the last sentence that says "The sum of the molecular masses of the reactants is 642.6 g/mol, so theoretically a cell can produce two faradays of charge (192,971 coulombs) from 642.6 g of reactants, or 83.4 ampere-hours per kilogram (or 13.9 ampere-hours per kilogram for a 12-volt battery)." It doesn't say anything about that being discharge rate - i.e. chemical reaction rate - dependent.
AFAIR doesn't your battery weigh about 25kg? (You can't calculate accurately from that because of the weight of the case, terminals etc, and the non-stoichiometric balance of reactants in a practical battery which will be +ve plate limited.)


Negative plate reactionPb(s) + HSO−4(aq) → PbSO4(s) + H+
(aq) + 2e⁻ Release of two conducting electrons gives lead electrode a net negative chargeAs electrons accumulate they create an electric field which attracts hydrogen ions and repels sulfate ions, leading to a double-layer near the surface. The hydrogen ions screen the charged electrode from the solution which limits further reactions unless charge is allowed to flow out of electrode.
Positive plate reactionPbO
2(s) + HSO−
4(aq) + 3H+
(aq) + 2e⁻ → PbSO
4(s) + 2H
2O(l)The total reaction can be written as
Pb(s) + PbO
2(s) + 2H
2SO
4(aq) → 2PbSO
4(s) + 2H
2O(l)

I haven't checked the accuracy of the figures in the calculation you show but they don't appear wildly innacurate so I've just used them. It's been a while since I worked as a research chemist and that was organic chemistry. However, I still remember some basic inorganic stuff.

You are describing theoretical capacity but stating it doesn't mention reduction due to change in discharge rate. That's entirely expected and doesn't preclude useable Ah changing under different conditions.

I accept that there's a theoretical limit but wouldn't expect a real battery to approach it. Many things will act against that. I'd imagine that plate thickness would essentially "hide" some of the active mass making it unavailable for reaction in a reasonable timescale. Electrolyte stratification probably takes place within a plate and reduces the reaction rate (one of several reasons). I guess that high discharge rates exceed the rate at which diffusion attempts to restore the balance. Hence the apparent increase in capacity at low discharge rates.

T105 weighs in at 28kgs casing and internal support structures are probably relatively light (polypropylene or similar).
So could be around 25kgs for the active mass in plates and that's 50kg for 2 x T105s acting as a 225Ah (C20) 12V battery.

You get close to 700Ah theoretical capacity (50 x 13.9Ah/kg).

I wouldn't expect the stoichiometric balance to be wildly off. It's never exact in real life and chemists usually do it deliberately for a variety of reasons. Even changing the balance so that max. reaction possible was 80% and knocking off another 10% from estimated active mass would still give >500Ah as theoretical capacity.

My experience would have led me to assume that 250Ah at 100 hour rate was well below the theoretical limit. This would appear to be the case.

When you said capacity didn't change I was thinking about useable capacity. However, I now believe that you have clarified that you mean theoretical capacity won't change.

This means:
250Ah (100 hour rate) is probably between 36-50% Th.
225Ah (20 hour rate) is probably between 32-45% Th.

It doesn't seem to be unreasonable for Trojan to quote 250Ah capacity under the stated conditions.

 

[pcatterall]

Lots of interesting theory but in the OPs situation, especially as this was going to be a regular occurrence, I would have done a well controlled test over just one or the 2 days and determined exactly what my batteries state was before and after. If I had used more juice than I was happy with then ( in future) I would either reduce consumption or install another battery.
Just one instance of going beyond the recommended value is not going to be a big issue and he is equipped with solid evidence for the future.

 

[lw395]

Clearly you cannot convert the whole of the lead plate to either oxide or sulphate, you need some plate left to function as a an electrical contact.

Conventionally the capacity is quoted at the 20hr rate, or the 5hr rate in the case of starter batteries. I think that is dictated by IEC and/or SAE standards.

 

[simonfraser]

Lots of interesting theory but in the OPs situation, especially as this was going to be a regular occurrence, I would have done a well controlled test over just one or the 2 days and determined exactly what my batteries state was before and after. If I had used more juice than I was happy with then ( in future) I would either reduce consumption or install another battery.
Just one instance of going beyond the recommended value is not going to be a big issue and he is equipped with solid evidence for the future.

Good point, i remote start the D2 so the boat is warm when i get there.
Would have to switch it all off and let it settle at either end of the test.

Having read all the info and looked at the graphs, i am going to believe the data from Trojan and risk it with two batteries.
Could buy a battery monitor, but that pays for another Trojan.

 

[pcatterall]

I get good use from my NASA monitor it gives me the confidence that perhaps just having another battery may not.
having said that I still do a real life test every 6months by running the batteries down to 50% against a known load.

 

[Plevier]

AIUI, there is inefficiency in discharge too, so the capacity you get out varies according to the rate at which you discharge it.
* viewed from outside the battery, yes; but from the aspect of conversion of active material, no

I get your point that if you take out say 50Ah at 50A, the end voltage will be lower than if you take out 50Ah at 5A. Because the 50A will be causing a significant drop from resistance etc. But if you let the battery settle for a day and then drew 5A in both cases, the battery which had be drained at 50A would generally have a lower voltage than the other.
* if you have taken exactly 50Ah out of each, then after the recovery time, they will be the same voltage.

At the other end of the scale, there is self discharge, so draining a bettery slowly can be inefficient too.
* yes but at typically 1-2% per week, not significant in the context of this discussion

I believe some batteries, drawing a small current can increase the self discharge. I don't know if that's true of lead acid.
* Don't know, never heard that.

.

 

[Plevier]

Let’s look at the AH capacities of a common solar battery, the Trojan L16 6 Volt battery. According to Trojan, the ratings are as follows:
Battery "C" Rating Battery Capacity Load Available Energy
C5 HOURS 303 AH 60.6 AMPS 1.82 KWH
C10 HOURS 340 AH 34.0 AMPS 2.04 KWH
C20 HOURS 370 AH 18.5 AMPS 2.22 KWH
C100 HOURS 411 AH 4.11 AMPS 2.47 KWH

From:
http://solarhomestead.com/battery-amp-hour-ratings/


I thought this was interesting too:
https://www.researchgate.net/deref/...cimages/7427/Lead_Acid_Battery_Efficiency.pdf

As a battery gets to higher charge levels, the charging efficiency plummets.
"These tests indicate that from zero SOC to 84% SOC the
average overall battery charging efficiency is 91%, and
that the incremental battery charging efficiency from 79%
to 84% is only 55%. "

So having batteries bigger than you need can be very inefficient.

Those kWh ratings are approximate and optimistic - they are calculated at 6V.

Yes charging efficiency tends to 0% as SoC tends to 100% but it's not relevant to this discussion.

Interesting point though about a big battery for a job being inefficient as regards charging energy consumption, I'd never thought of that. Have to balance it against what cycle life you need to have from a reliability point of view.

 

[lw395]

Those kWh ratings are approximate and optimistic - they are calculated at 6V.

Yes charging efficiency tends to 0% as SoC tends to 100% but it's not relevant to this discussion.

Interesting point though about a big battery for a job being inefficient as regards charging energy consumption, I'd never thought of that. Have to balance it against what cycle life you need to have from a reliability point of view.

Indeed, the available energy is just the Ah x 6V. So the real energy is going to have a bigger spread as the terminal volts are lower at higher currents.
I don't think efficiency as such is really a goal for many of us, but the way it changes must make calibration of battery monitors hard.

 

[Plevier]

Lots of interesting theory but in the OPs situation, especially as this was going to be a regular occurrence, I would have done a well controlled test over just one or the 2 days and determined exactly what my batteries state was before and after. If I had used more juice than I was happy with then ( in future) I would either reduce consumption or install another battery.
Just one instance of going beyond the recommended value is not going to be a big issue and he is equipped with solid evidence for the future.

Yep!

 

[Plevier]

Clearly you cannot convert the whole of the lead plate to either oxide or sulphate, you need some plate left to function as a an electrical contact.

Conventionally the capacity is quoted at the 20hr rate, or the 5hr rate in the case of starter batteries. I think that is dictated by IEC and/or SAE standards.

Lead dioxide is actually pretty conductive. The main current paths are of course through the lead alloy grid framework of the plates. That does not convert to oxide, except a thin surface layer. The active material is the porous paste which is pressed into the grid in manufacture (and then baked).

There are more arbitrary capacity standards with differing times, voltages and temperatures than you can shake a stick at, to use a silly expression! To add to it, the trend in automotive is to use reserve capacity - that's the number of minutes for which the battery will supply a specified current, I think it's 20A but that's just from memory.

 

[Plevier]

Indeed, the available energy is just the Ah x 6V. So the real energy is going to have a bigger spread as the terminal volts are lower at higher currents.
.

More important than that, it's because the terminal voltage will be drop during discharge, possibly to 1.7vpc in a standby UPS system which is one of the few truly constant power loads. That means a commensurate increase in current. You can go that low because it is not designed to do it repeatedly, it's an emergency standby.

 

[Plevier]

I haven't checked the accuracy of the figures in the calculation you show but they don't appear wildly innacurate so I've just used them. It's been a while since I worked as a research chemist and that was organic chemistry. However, I still remember some basic inorganic stuff.

You are describing theoretical capacity but stating it doesn't mention reduction due to change in discharge rate. That's entirely expected and doesn't preclude useable Ah changing under different conditions.

I accept that there's a theoretical limit but wouldn't expect a real battery to approach it. Many things will act against that. I'd imagine that plate thickness would essentially "hide" some of the active mass making it unavailable for reaction in a reasonable timescale. Electrolyte stratification probably takes place within a plate and reduces the reaction rate (one of several reasons). I guess that high discharge rates exceed the rate at which diffusion attempts to restore the balance. Hence the apparent increase in capacity at low discharge rates.

T105 weighs in at 28kgs casing and internal support structures are probably relatively light (polypropylene or similar).
So could be around 25kgs for the active mass in plates and that's 50kg for 2 x T105s acting as a 225Ah (C20) 12V battery.

You get close to 700Ah theoretical capacity (50 x 13.9Ah/kg).

I wouldn't expect the stoichiometric balance to be wildly off. It's never exact in real life and chemists usually do it deliberately for a variety of reasons. Even changing the balance so that max. reaction possible was 80% and knocking off another 10% from estimated active mass would still give >500Ah as theoretical capacity.

My experience would have led me to assume that 250Ah at 100 hour rate was well below the theoretical limit. This would appear to be the case.

When you said capacity didn't change I was thinking about useable capacity. However, I now believe that you have clarified that you mean theoretical capacity won't change.

This means:
250Ah (100 hour rate) is probably between 36-50% Th.
225Ah (20 hour rate) is probably between 32-45% Th.

It doesn't seem to be unreasonable for Trojan to quote 250Ah capacity under the stated conditions.

Oh ****** I wrote a reply where's it gone!!! What did I say?

Didn't realise you were a chemist, sorry!

I think we have come full circle.
I don't disbelieve the claimed capacities at different rates, this has all been about the depth of discharge it means and I think you are accepting the point.
In your words, it's an increase in APPARENT capacity.
My point is that for depth of discharge calculation for cycle life purposes, you cannot compare with this increased apparent capacity. You must continue to compare with NOMINAL capacity which is some proportion of the theoretical (your word) or chemical (my word) capacity.

Oh yes, stoichiometry - there's typically 10-15% excess of negative plate material and a somewhat larger excess of electrolyte as you don't want the s.g. to go too low during discharge.
OCV = (s.g + 0.85)VPC at 25 deg C pretty accurately over the normal working range.

 

[Mistroma]

Oh ****** I wrote a reply where's it gone!!! What did I say?

Didn't realise you were a chemist, sorry!

I think we have come full circle.
I don't disbelieve the claimed capacities at different rates, this has all been about the depth of discharge it means and I think you are accepting the point.
In your words, it's an increase in APPARENT capacity.
My point is that for depth of discharge calculation for cycle life purposes, you cannot compare with this increased apparent capacity. You must continue to compare with NOMINAL capacity which is some proportion of the theoretical (your word) or chemical (my word) capacity.

Oh yes, stoichiometry - there's typically 10-15% excess of negative plate material and a somewhat larger excess of electrolyte as you don't want the s.g. to go too low during discharge.
OCV = (s.g + 0.85)VPC at 25 deg C pretty accurately over the normal working range.

Not so much a chemist as an ex. chemist (but not yet ex. in Python terminology). Busy day so barely time to get near a computer till now. At least my car has a new timing belt in time for an upcoming 900 mile round trip and I hopefully won't have to spend hours shovelling snow tomorrow.

I imagine that slow discharge allows diffusion to keep up with the need to move reactive species deep into the plates. I mean plate area is fixed and not very large, so variation in capacity ought to come from conversion of material deeper within the plate. I know that there will be a lot of things going to cause migration of various species and at least one would be going "uphill" electrically speaking (and that would favour surface discharge). However, just generally using migration without specifics probably covers most of what's happening.

Assuming 2 batteries discharged to 50% based on C20 and C100 rate:
Conversion of deeper plate material should take more "effort/time" to reverse than at the surface. Fully recharging a slowly discharged battery will involve a higher proportion of active mass deep inside a plate. Nothing new here, after all more Ah taken out. But the slow discharge should result in a more even change in plate material in relation to distance from plate surface. Not the same at surface and deep inside the plate, just less of a gradient moving away from the plate's surface.

In OPs case he probably starts at full charge on Sat. morning and returns to mains charging on Sun. afternoon. Batteries will probably spend little time at 50% and most at 100%. I doubt it makes a huge difference whether he takes 50% of 225Ah or 250Ah in terms of longevity.

My case is different (and selfishly, the one I think about more). I spend almost all of my time at anchor so can't get back to 100% for majority of the time. I try to minimise current draw and it mostly sits around 6A when fridge is cycling but averages 4. LED lights and running laptops on internal batteries at night helps. I'm certainly in the region where Trojan quote 250Ah capacity so running batteries to 50% would be more damaging than normal as you say it's actually less than 50% (because I should compare it with 225A, not 250Ah).

The problem I have at present is the unexplained drop in accuracy of my Smartgauge. I thought that it might be explained in part by your contention that 50% of C100 Ah figure would be more than 50% discharged as I should use C20 figure for actual capacity. Problem with batteries is that nothing is fixed, measurements are not usually very accurate and you have no means of assessing how much lifespan is left.

I'll look at my historical data again anyway to see if slow discharge is part of the reason for the anomaly. I improved fridge insulation slightly, altered solar charging setup and had some electrical problems this year. Smartgauge usually works well (by magic, aka high freq. pulses) and suddenly didn't. All electrical connections were checked and Merlin are bench-testing my instrument. I'm hopeful that I'll get to the bottom of it. I'll ask what they think about possibility of reduction in my usage leading to Smartgauge reporting my batteries had a much lower SoC than SG measurments suggest.

On the bright side, slow discharge must have some benefits which offset reaction of more active mass. I think a traction battery is mainly degraded by loss of cohesion of active mass on the plate and sulfation. I can see high discharge rates degrading cohesion more rapidly than low discharge rate, so it might not all be bad news (it's all relative anyway ).

 

[WightMistress]

IMHO you can never have too many batteries. We have spent the last couple of winters re-kitting Wight Mistress (F34) with as many LED lights - now only a couple of filament lamps left to change. We have set up our fridge with it's own AGM battery, keel cooler and dedicated solar panel (completely separate from the domestic circuit.) We no longer have a built in mains battery charger, and we currently have 2 x 100AH wet cells as a domestic bank.
They appear to be doing really well, but we have decided to install our spare pair of 100AH batteries and double the domestic bank size, and go for a second solar panel just for the domestic bank. Our heavy loads include a laptop for navigation, plus all the other navaids, and our Wallas diesel hob takes a bit of juice when the hot plug is running on start up. I did a spreadsheet of our electricity budget and if everything is burning and turning, my estimate of current comes out in the order of 70 amps. Of course, when sailing we reduce load as far as possible, and at night that runs at around 23 amps with the laptop, AIS transponder and all the gadgets on. We tend to use the Hydrovane rather than the electric Cetrek autopilot when under sail. Current capacity 50% of bank size = max 100 ah, so running at 20amps is really not big enough. This said, we get across the Channel without problems, only having to run under power for short periods. With the addition of the extra 2 x 100AH, the project will include shunts and current monitoring, plus a few hours or learning how to drive it properly.
Hope this is useful.

 

[Plevier]

Not so much a chemist as an ex. chemist (but not yet ex. in Python terminology). Busy day so barely time to get near a computer till now. At least my car has a new timing belt in time for an upcoming 900 mile round trip and I hopefully won't have to spend hours shovelling snow tomorrow.

I imagine that slow discharge allows diffusion to keep up with the need to move reactive species deep into the plates. I mean plate area is fixed and not very large, so variation in capacity ought to come from conversion of material deeper within the plate. I know that there will be a lot of things going to cause migration of various species and at least one would be going "uphill" electrically speaking (and that would favour surface discharge). However, just generally using migration without specifics probably covers most of what's happening.

Assuming 2 batteries discharged to 50% based on C20 and C100 rate:
Conversion of deeper plate material should take more "effort/time" to reverse than at the surface. Fully recharging a slowly discharged battery will involve a higher proportion of active mass deep inside a plate. Nothing new here, after all more Ah taken out. But the slow discharge should result in a more even change in plate material in relation to distance from plate surface. Not the same at surface and deep inside the plate, just less of a gradient moving away from the plate's surface.

* What you say is certainly correct but I don't think it's a significant factor in discharges of more than an hour, maybe less. Don't forget normal leisure battery plates will be 1-2mm thick, I think the T-105 plates are between 3 & 4mm (possibly 1/8"), I expect the 12V Trojans are thinner. Diffusion depth only needs to be half that of course.

In OPs case he probably starts at full charge on Sat. morning and returns to mains charging on Sun. afternoon. Batteries will probably spend little time at 50% and most at 100%. I doubt it makes a huge difference whether he takes 50% of 225Ah or 250Ah in terms of longevity.

* Cycle life figures aren't very consistent anyway.

My case is different (and selfishly, the one I think about more). I spend almost all of my time at anchor so can't get back to 100% for majority of the time.

* Imust point out - batteries don't like that!

I try to minimise current draw and it mostly sits around 6A when fridge is cycling but averages 4. LED lights and running laptops on internal batteries at night helps. I'm certainly in the region where Trojan quote 250Ah capacity so running batteries to 50% would be more damaging than normal as you say it's actually less than 50% (because I should compare it with 225A, not 250Ah).

The problem I have at present is the unexplained drop in accuracy of my Smartgauge. I thought that it might be explained in part by your contention that 50% of C100 Ah figure would be more than 50% discharged as I should use C20 figure for actual capacity. Problem with batteries is that nothing is fixed, measurements are not usually very accurate and you have no means of assessing how much lifespan is left.

* How is your Smartgauge misbehaving, and how do you know it is? Sure you aren't shooting the messenger?

I'll look at my historical data again anyway to see if slow discharge is part of the reason for the anomaly. I improved fridge insulation slightly, altered solar charging setup and had some electrical problems this year. Smartgauge usually works well (by magic, aka high freq. pulses) and suddenly didn't. All electrical connections were checked and Merlin are bench-testing my instrument. I'm hopeful that I'll get to the bottom of it. I'll ask what they think about possibility of reduction in my usage leading to Smartgauge reporting my batteries had a much lower SoC than SG measurments suggest.

* On a battery in good condition, OCV is a good indicator of SoC. On a battery in poor condition, it can be very misleading.

* I don't know much about Smartgauge and they are rather secretive. My guess is that somehow it must look at (dV/dt) and instantaneous voltage, those 2 parameters define a unique point on the family of discharge curves for a battery. I don't know how it builds up its library of curves. Does it take a long time to adapt to a new battery? How much deterioration in battery performance can it cope with?

On the bright side, slow discharge must have some benefits which offset reaction of more active mass. I think a traction battery is mainly degraded by loss of cohesion of active mass on the plate and sulfation.
* loss of contact with grid, corrosion of grid (worse with lead calcium than with lead antimony!), cracking within body of paste, loss of paste from surface of plates, sulphation shouldn't be a factor unless the charge regime is poor

I can see high discharge rates degrading cohesion more rapidly than low discharge rate, so it might not all be bad news (it's all relative anyway ).
* not sure about that. Your previous argument is that at high rate, reaction - and therefore the damaging volumetric changes - will be more superficial. That will affect the grid/paste interface less! I believe grid/paste interface suffers badly in very deep discharges.

.

 

[Mistroma]

Yes, nothing I'm certain about. However, my thinking was that more aggressive changes were concentrated nearer to surface with high discharge and lower reaction rates deeper inside the plate. I was thinking that the surface layers might progressively lose material due to loss of cohesion. My assumption was that loss of material takes place mainly at the surface. I know that aggressive charging will speed up loss of active mass from plate surface.

Deep discharges obviously cause damage due to sulfation but I wasn't certain of the impact on plate cohesion unless left discharged. I can see slower large crystal growth having adverse affect on a battery left discharged and even cause physical plate damage.