There are no forum members that confuse A and AH.

[noelex 77]

Originally Posted by wholybee

This is incorrect.

You are confusing the energy that is entering the battery and the energy that can be recovered from the battery.

If we put in one amp at 14.7v for one hour (14.7 Whrs) and leave the battery to sit overnight, we cannot draw 14.7 Whrs from the battery and end up at the same state of charge. When we try to draw energy out of the battery, even if a lead acid battery has 100% SOC the voltage will have dropped to around 12.7v. If we want to draw out 14.7 Whrs we would need a current of 1.16A for one hour. This does not work and is why battery monitors count AHrs not Whrs.

Try designing a battery monitor that counts Whrs in and Whrs out. This would be easily done, but it will not work well.

On the other hand if we want to calculate the energy that a windlass will use, as you point out, Whrs are better than Ahrs. This is why both measurements are useful.

[wholybee]

Originally Posted by noelex 77

You are confusing the energy that is entering the battery and the energy that can be recovered from the battery.

If we put in one amp at 14.7v for one hour (14.7 Whrs) and leave the battery to sit overnight, we cannot draw 14.7 Whrs from the battery and end up at the same state of charge. When we try to draw energy out of the battery, even if a lead acid battery has 100% SOC the voltage will have dropped to around 12.7v. If we want to draw out 14.7 Whrs we would need a current of 1.16A for one hour. This does not work and is why battery monitors count AHrs not Whrs.

Try designing a battery monitor that counts Whrs in and Whrs out. This would be easily done, but it will not work well.

On the other hand if we want to calculate the energy that a windlass will use, as you point out, Whrs are much better than Ahrs. This is why both measurements are useful.

This is the result of efficiency and surface charge. If you charge a battery at 14.7 V at 1 Amp for 1 hour, you did not put 14.7 Ah into the battery. You actually put in somewhat less. And when the battery voltage drops overnight, that is not energy lost, it is the surface charge dissipating. The error you describe from your test would be the same with Ah and Wh, because the ONLY difference between the 2 is that Ah doesn't include Voltage.

Edit:
Also, the test isn't fair at 14.7V, because the battery is fully charged at that point, and additional energy is lost due to overcharging/heating.

[noelex 77]

Originally Posted by wholybee

This is the result of efficiency and surface charge. If you charge a battery at 14.7 V at 1 Amp for 1 hour, you did not put 14.7 Ah into the battery. You actually put in somewhat less. And when the battery voltage drops overnight, that is not energy lost, it is the surface charge dissipating. The error you describe from your test would be the same with Ah and Wh, because the ONLY difference between the 2 is that Ah doesn't include Voltage.

I think you mean 14.7 Whrs. This is exactly what a meter measuring the energy entering the battery would record. A meter measuring Ahr would record 1 Ahr.

We can at a later stage withdraw 1 Ahr from the battery and return close to the same SOC (a small factor for battery inefficiency needs to factored in), but we cannot withdraw 14.7 Whrs without being at a significantly lower SOC. Ahrs in closely matches Ahrs out. Whrs in does not closely match Whrs out.

The error is not the same for Ahrs and Whrs. Ahrs are significantly more accurate in this application which is why they are used.

This is what Jedi means when he talks of counting Coulombs to keep track of SOC.

Originally Posted by s/v Jedi

With lithium we can almost take it as an ideal storage medium but for SOC we still need to count Coulombs.

One Coulomb is one amp for one second. So in layman’s terms counting Coulombs is counting Ahrs, just over a different time frame (an hour rather than a second).

Ahrs are better than Whrs for some applications, such as tracking battery SOC.

[catalystcat]

Originally Posted by s/v Jedi

As battery technology evolves, the difference becomes less. With lithium we can almost take it as an ideal storage medium but for SOC we still need to count Coulombs.

I have the Peukert’s factor at 1.00 though

If that's at all temperatures (or even just a few above absolute 0), I'd like to buy one, especially if the next significant digit is also 0....

Speaking more generally, given the multiple dependencies here that affect measurement accuracy - e.g., calculating energy from Ah by assuming a constant voltage; Peukert factor difference over time and temp and chemistry; voltage measurement at rest and under load - I am surprised this conversation hasn't devolved into a discussion of precision vs accuracy in measuring/data.

On a practical note, I am also surprised that battery and current monitor systems don't seem to build a more sophisticated model of energy status as a function of how much energy is being used, stored, and delivered and the dependencies on things like peak power rate vs duty cycle vs partial discharge vs temperature vs cable losses vs ... etc.

The data is available to monitor current, voltage, temperature, age, charge cycle history, etc. It's not all that complicated to track this data over a wide range of operating conditions and integrate that in a model of how battery capacity ages over time, and how it depends on things like those mentioned above.

In fact, people kind of do a crude mental model already in a way when they look at voltage, and adjust their power available estimates based on how "aged" their batteries are, on how cold it is, and some history.

There are a lot of baby steps along this path - e.g., charts of battery capacity dropoff over time, Peukert factor inputs to inverter chargers, temperature monitoring of batteries, etc. But, I haven't seen any attempts to really integrate all or most of this available information into a model. Note that with the evolution of machine learning systems (I know the trendy usage is AI, but, ML really seems more appropriate here), each system can probably adapt and refine the model to individual boats, eventually.



It would be a really interesting feature for some company or user group to implement at least a simpler version of this kind of thing.

BTW - There are a lot of interesting side benefits that might emerge, like battery failure/lifetime predictions based on model drift or specific data patterns.
(One of my favorite analogies is an old predictive model of hard drive failures in datacenters that used acoustic data. Drives that were on their way to failure had a recognizable shift in their acoustic patterns. In pre-nonvolatile memory days that saved many millions of dollars in downtime costs at datacenters.)

[s/v Jedi]

Originally Posted by wholybee

This is incorrect.

Ah = Amperage x hours
Wh = Watts x hours
W = Voltage x Amperage
Wh = Voltage x Amperage x hours

The ONLY difference in accuracy between the two units, is that Ah doesn't include Voltage. In your example, at the lower voltage, there really *is* less stored energy, because the voltage is lower. The stored energy is *not* close to equal. It might be "1 Ah" at both high state of charge and low state of charge. But if you are running a windlass for one hour at 1 Amp at both times at the lower voltage less work is done(the windless runs slower). So, the lower state of charge has less stored energy, even if you figure 1 Ah at both charge levels.

For stored energy, you are correct. But not for state of charge of the battery. If you put 1Ah into a 100Ah battery at 12.2V then that is 1% and when you put 1Ah into a battery at 14.4V then that is 1% as well.

That battery, if LA would hold 1.2kWh energy because the nominal cell voltage is 2V and we have 6 cells in series.

[s/v Jedi]

Originally Posted by noelex 77

I think you mean 14.7 Whrs. This is exactly what a meter measuring the energy entering the battery would record. A meter measuring Ahr would record 1 Ahr.

We can at a later stage withdraw 1 Ahr from the battery and return close to the same SOC (a small factor for battery inefficiency needs to factored in), but we cannot withdraw 14.7 Whrs without being at a significantly lower SOC. Ahrs in closely matches Ahrs out. Whrs in does not closely match Whrs out.

The error is not the same for Ahrs and Whrs. Ahrs are significantly more accurate in this application which is why they are used.

This is what Jedi means when he talks of count Coulombs to keep track of SOC.

One Coulomb is one amp for one second. So in layman’s terms counting Coulombs is counting Ahrs, just over a different time frame (an hour rather than a second).

Ahrs are better than Whrs for some applications, such as tracking battery SOC.

Exactly, we agree

[s/v Jedi]

Originally Posted by catalystcat

If that's at all temperatures (or even just a few above absolute 0), I'd like to buy one, especially if the next significant digit is also 0....

Speaking more generally, given the multiple dependencies here that affect measurement accuracy - e.g., calculating energy from Ah by assuming a constant voltage; Peukert factor difference over time and temp and chemistry; voltage measurement at rest and under load - I am surprised this conversation hasn't devolved into a discussion of precision vs accuracy in measuring/data.

On a practical note, I am also surprised that battery and current monitor systems don't seem to build a more sophisticated model of energy status as a function of how much energy is being used, stored, and delivered and the dependencies on things like peak power rate vs duty cycle vs partial discharge vs temperature vs cable losses vs ... etc.

The data is available to monitor current, voltage, temperature, age, charge cycle history, etc. It's not all that complicated to track this data over a wide range of operating conditions and integrate that in a model of how battery capacity ages over time, and how it depends on things like those mentioned above.

In fact, people kind of do a crude mental model already in a way when they look at voltage, and adjust their power available estimates based on how "aged" their batteries are, on how cold it is, and some history.

There are a lot of baby steps along this path - e.g., charts of battery capacity dropoff over time, Peukert factor inputs to inverter chargers, temperature monitoring of batteries, etc. But, I haven't seen any attempts to really integrate all or most of this available information into a model. Note that with the evolution of machine learning systems (I know the trendy usage is AI, but, ML really seems more appropriate here), each system can probably adapt and refine the model to individual boats, eventually.

It would be a really interesting feature for some company or user group to implement at least a simpler version of this kind of thing.

BTW - There are a lot of interesting side benefits that might emerge, like battery failure/lifetime predictions based on model drift or specific data patterns.
(One of my favorite analogies is an old predictive model of hard drive failures in datacenters that used acoustic data. Drives that were on their way to failure had a recognizable shift in their acoustic patterns. In pre-nonvolatile memory days that saved many millions of dollars in downtime costs at datacenters.)

But there is. Check out the Vicron Cerbo GX and the free portal website that does all the collecting of statistics to keep track of that. It’s great, you can exactly see what’s going on aboard, if the current weather supports your energy needs etc.

[Sand crab]

So, are all amp hours created equal? We were having this debate about some cordless tools with different voltages. Does a 12 volt 4.0 ah battery have the same power as an 18v 4.0ah battery?

[Chotu]

Originally Posted by Sand crab

So, are all amp hours created equal? We were having this debate about some cordless tools with different voltages. Does a 12 volt 4.0 ah battery have the same power as an 18v 4.0ah battery?

They are not created equal.


Watts is the unit of power. Watts = volts x amps. The bigger the volts get for the same amount of amps, the bigger the watts are.

So, higher voltage 4ah batteries have more stored energy.

[SailingHarmonie]

Originally Posted by Sand crab

So, are all amp hours created equal? We were having this debate about some cordless tools with different voltages. Does a 12 volt 4.0 ah battery have the same power as an 18v 4.0ah battery?

To keep with the theme of this thread I should chastise you for an incorrect use of the term "power" when you ACTUALLY mean energy...

But NO. Just to make the math easier, a 100 Amp-hour 12 volt battery has half as much energy stored as a 100 Amp-hour 24 Volt battery.

[s/v Jedi]

Correct. Now consider how much of that energy is usable.

[catalystcat]

Originally Posted by s/v Jedi

But there is. Check out the Vicron Cerbo GX and the free portal website that does all the collecting of statistics to keep track of that. It’s great, you can exactly see what’s going on aboard, if the current weather supports your energy needs etc.

I spend a lot of time on Victron Connect and VRM looking at the data from my Cerbo. In fact, I used it on my phone yesterday to check in on my batteries in Grenada while I am just down the road from you in Brevard County.

It's very useful and is a good repository of data from my system and has some nice UI that shows things like charge history, power usage, real time charging data, etc. It is a very good start toward what I was suggesting. It knows how many Ah I have and uses simple algorithms to report a guess about how much is remaining.

What it lacks is a good computer model, especially one that actually uses the available data to evolve itself. They have access to a huge amount of data from my system and others that can be used to improve accuracy. For example, the Cerbo uses an input Peukert factor, efficiency, and a tail current, and does not refine those over time.

The data is there to build a good model of battery aging/energy capacity as a function of loads, cycles, temperatures, battery models, balancing intervals, and more. This is the kind of thing that ML can do very well given enough data and, via Victron Connect, they have a ton of data.

[Baronkrak]

Originally Posted by wholybee

W = Voltage x Amperage

W = Voltage x Amperage
Is incorrect.

W = V x A x p.f.
VA = V x A

[JBP]

Originally Posted by s/v Jedi

As battery technology evolves, the difference becomes less. With lithium we can almost take it as an ideal storage medium but for SOC we still need to count Coulombs.

I have the Peukert’s factor at 1.00 though

I think Noelex's point was that by measuring Amp hours and ignoring the charge/discharge voltages we introduce an error that nearly counter balances the losses of a lead acid battery, eliminating the need to calculate the loss if we were to more accurately measure total energy as coulombs or Wh. Although lithium batteries are much more efficient, the difference in charge vs discharge voltage is also much less, and thus measuring in Ah and ignoring actual accurate energy measurement again results in counter balanced errors that very conveniently allow us to use Ah to pretend we have perfect batteries and ignore losses. Of course measuring in Ah does nothing to counter balance Peukert's law, but conveniently LFP batteries don't suffer much from Peukert's law at the current drains normally seen on boats.

[s/v Jedi]

Originally Posted by JBP

I think Noelex's point was that by measuring Amp hours and ignoring the charge/discharge voltages we introduce an error that nearly counter balances the losses of a lead acid battery, eliminating the need to calculate the loss if we were to more accurately measure total energy as coulombs or Wh. Although lithium batteries are much more efficient, the difference in charge vs discharge voltage is also much less, and thus measuring in Ah and ignoring actual accurate energy measurement again results in counter balanced errors that very conveniently allow us to use Ah to pretend we have perfect batteries and ignore losses. Of course measuring in Ah does nothing to counter balance Peukert's law, but conveniently LFP batteries don't suffer much from Peukert's law at the current drains normally seen on boats.

Yes, sorry, with “counting Coulombs” I mean Ah, it’s the same. I’m so used to talk about Coulomb counting that I didn’t realize this (yet another) unit wasn’t discussed yet, complicating this discussion