Measuring Charge Efficiency
 

There has been some discussion of replacing a lead-acid car battery with super capacitors and/or LiPoFe4 batteries to save weight and possibly eliminate the need to periodically replace the battery. Some sources place charging efficiency for lead-acid as low as 50% when charging near full capacity. Sources also claim charge efficiency as high as 90% for LiPoFe4 at near full charge, and >95% efficiency for charging capacitors.

Not only is there charge efficiency to consider, but also discharge.

I'm looking for a way to compare charge/discharge efficiency. My idea is to measure cumulative Wh or Ah output from the alternator for a given distance or engine running time, and compare the results to battery alternatives.

The problem is controlling for the many variables in electrical use. If the test period wasn't sufficiently long enough, a single trip at night running the headlights would significantly skew the results.

1. What device would I use to measure the cumulative Wh or Ah? I'm thinking the Watt's Up would do the trick, but don't know how it would work, or how to protect it from the large discharge of starting a motor.

2. How should I control my variables to get the most accurate efficiency results? Ideally I would measure the amount of power going into the battery, and the amount going out. Any idea how to accomplish this? Perhaps bench charging/discharging is a better way to control the variables, but I'm very curious of what the results would be in actual vehicle use.

I realize the impact on fuel efficiency is minuscule, but my curiosity must be satisfied!

My testing would be as follows:
1) charge both batteries overnight, and make sure voltage is the same.
2) start car on a 3rd battery
3) install battery of choice
4) connect ammeter in series
5) let car idle with typical accessory load for one hour.

There are still variables,such as temp, and self discharge rates, but this eliminates most. You would need a guino or something that is measuring injector pulses for your reading.

Am I correct in assuming that when a car is running, most of the electrical loads are fed directly from the alternator, eliminating most of the charge/discharge inefficiency of the battery?

If that is true, then charge/discharge efficiency really only comes into play when an electrical load exceeds alternator output.

Is this main place where gains are to be had, once the engine is running, battery recharged, accessories running off the alternator... Is the battery just sitting there suckin up the extra voltage just turning it into heat?
Last I read up on newer gm's charging systems, when the charging system is in fuel economy mode, the output voltage is targeted at what it thinks the open circuit voltage of the battery is. Didn't give it much thought previously, but I guess gm's strategy is to try to leave the battery out of it as much as possible.

Also, any thoughts to if an AGM battery might outperform a lead acid in this respect on an older vehicle due to its slightly higher voltage?

calculated load
 

If you supply a battery with a constant 13.5v (or whatever voltage you typically see in your car while maintaining highway speeds) for an unlimited amount of time, what is the lowest the amperage the battery falls to?
I'm wondering if it could be a couple amps for a lead acid, just slightly less for an agm, but for capacitors or lifepo4 batteries that 13.5 wouldn't be near as high state of charge, would it be near 0?

I'm not sure I followed your whole train of thought, but a capacitor that is given a sustained 13.5v will take on charge very rapidly until it approaches 13.5v. As it gets ever closer to 13.5v, the current drops off until there is practically zero current flow. Therefore, a capacitor is very efficient in that it no longer consumes energy when it has reached equilibrium voltage with the alternator.

Am I on the right track with your question?

I'm curious if a lead acid battery dissipates energy even though it is already at full charge. If it does, I wonder how that compares to a LiFePO4 battery considering lithium chemistries require current flow to cease when full charge is reached.

I just got 2 ammeters with logging capability, so I can begin to compare how much energy goes into a battery vs how much goes back out.

The only problem is that the meters are rated at 100A, and measuring the starter would blow them. Starting is the biggest single drain on a battery, so it would be important to capture the data. Once a car is running, the battery rarely supplies power to anything, and when it does, it's for a brief moment.

Once LA batteries approach full charge their charge acceptance begins to drop. Modern "smart' alternators look for this drop in charge acceptance to determine the level of charge in the battery (when the alt. is running voltage is pretty much a constant 13.8v regardless of battery charge).

I do believe the excess current can only become heat, much like when you bench charge a LiFe and it gets warm as it gets full.

How would you quantify it? It seems more like something to do on a bench than in an engine bay, which would give you far greater control over variables. You could possibly spin up an alternator with a power drill, and add load via lights or 12v motors.

Sounds like you're on the same track. Watching amperage with a steady 13.5v might really be just a way to measure what the self discharge rate is at that voltage. A lead acid near 100% I would think have a significantly higher self discharge than one at 90%, I'm thinking that is where the oem's are trying to grab a little fuel economy from. A lifepo4 or capacitors are at a lower state of charge at the same voltage, so there might be very little self discharge. You wouldn't need to worry about stopping current flow to a lifepo4 because it wouldn't be fully charged anyway.

Unfortunately this looks like it would at best save about as much as having a taillight burned out.
What I'm thinking is that this would be the same as a float charge. Battery tender has some charts putting float charge at 50-100mA and chargetek says c/50-c/100 which I think would work out to slightly less for most batteries. So there's only about 5 to 13.5 watts on the table to try to save

Agreed. Charging and discharging efficiency is a miniscule drop in the overall efficiency of petrol vehicles. Charge efficiency only needs to be considered in systems where they are heavily used, such as EVs, alternator deletes, and hybrid traction batteries.

Regardless, I'm curious, and I don't see much information available from internet searches. My ultimate goal is to do an alternator delete while preserving as much convenience and longevity of the battery as possible.

Why not just test on the bench with a charger and a consumer?

I thought you were calling me a dummy load. :p

I brought my LiFePO4 and Supercap to work along with some resistors and meters and should be able to complete some tests. I'll start with 10% depth of discharge and then go down from there to 50% DoD and chart efficiency.

If helpful.
For A123 Flavor LiFePO4 20Ah cells.

Over a extended period of time (months) they have much much less self Discharge at any given temperature than more conventional PbA.
Link

Available discharge capacity is reduced at lower temperatures. Similar effect in PbA (but different magnitude).
Link

The complete cycle (Charge + Discharge) efficiency (Ah or Wh) reduces at higher rates (A or W).
Varying discharge rates Link

With no other changes (rate,temperature,etc) the cycle (Charge + Discharge) efficiency ( Ah or Wh ) reduces at high SoC/SoE.
Link
Link

The electrical resistance reduces at Higher SoC/SoE , but Increases at lower temperatures... Less resistance = Higher efficiency ... Agrees with above efficiency vs SoC/SoE measure.
Link

Peukert Effects of reduced Ah/Wh are almost entirely in the direction of the increased rate ... meaning if your Discharge rate increases but your charge rate does not ... less Ah/ Wh will be seen on the discharge part of the cycle ... and the opposite is also true ... if your charge rate increases but your discharge rate does not ... less Ah/Wh will be seen on the charge part of the cycle ... this effect is almost entirely on the CC portion of charge or discharge ... if a CV stage to charge or discharge is also included very little of this effect will actually be seen ... although, the time of the CV portion will greatly be effected by the CC rate portion.

• I saw less than 1% deviation on the charge (Ah or Wh) from deviation on the discharge rate from 10A to 100A.
• The end of CC portion moved ~5% (SoC/SoE) away from the bottom for discharge varying from 10A to 100A.
• I saw less ~0.5% deviation on the discharge (Ah or Wh) from deviations on the charge rate from 10A to 75A.
• the end of CC portion moved ~5% (SoC/SoE) away from the top for charge varying from 10A to 75A.

Taking what we know above combined with the OEM power vs SoC graph.
Link
We can also see a significant preference in allowable power pulses in the discharge direction ... Not just raw Watts ... but at the top SoC the Regen hits the 3.8v listed @ ~900W that's ~236Amps ... but on the opposite end the bottom discharge hits the 1.6v listed @~800W that's ~500Amps... which is also interesting considering the deviation we saw above in the resistance changes at those different ends of the SoC band.

- - - - -
I used the following:

• 3 PowerLab8's connected together (up to 120A rates charge or discharge combined).
• VA18B multi-Meter
• Instek GPD-2303S Power Supply
• 36 NiMH 6 cell Sticks from HEV Civic for high power for the PL8s.

~32MB compressed raw data from years of testing if desired

- - - - - -

Forgot to add ... these cells prefer to be kept bellow ~120F , for best life.
Link
Link