2013 EnerSys Genesis NP18-12B KOP Battery/FIRST Choice Weak?

[Chris_Elston]

I was just wondering something.

Our electrical team has gotten around to connecting and charging up the 2 Genesis batteries we got from FIRST Choice and the 1 battery we received from the KOP.

Have you load tested these batteries?

Generally we will cycle out batteries that fail our team specification. Our tester is a standard low-man's automotive 100-AMP load tester, that will load a battery for 10 seconds with a heat resistor and check for weakness using a CCA (cold crank amps) scale, we don't have a computerized load tester yet...this might be the year we purchase one...Typically the automotive 12V standard is 500 CCA for cars.

There will be someone that will probably argue CCA is not a good way to determine battery capacity. But we are establishing a base line with the automotive tester across our batteries and the new ones are the lowest numbers we have ever seen.

Anyway, after charging up the new Genesis batteries from 2013, they are posting about CCA 300 AMPs. Our team spec is no less than CCA 500 AMPs. Generally when we get new batteries, we see about CCA 700 AMPs to CCA 800 AMPs. We have seen this in both EnerSys and MK Batteries in the years past.

However this year...it's different...all three new 2013 batteries...CCA 300 AMPs...

I reference CCA because that is the tester we have, an automotive one.

While we appreciate the donation and are most thankful, we may not use these batteries on the robot this year. We might just repurchase some others. Unless there would be someone that could give us some insight into why we are only seeing CCA 300 AMPs on the 2013 batteries and maybe recommend a better more proven method to measuring battery capacity. Traditionally, we have seen the robot voltage drop below 11 Volts at the end of a 2.5 minute match with a CCA less than 500 AMP battery. Anything higher than that and the battery is generally good enough to even play a second match back-to-back. (Pending how many motors we have that particular year...)

So my questions are:

1. Do you load test batteries?
2. Did you load test the EnerSys Genesis NP18-12B 2013 KOP and FIRST Choice Batteries?
3. What was your results?
4. Can you confirm if your batteries meet your team spec?

Being that 2 batteries came from one source, and 1 battery came from another source...I am wondering if we are an isolated case or not.

[Hugh Meyer]

Chris,

We have started load testing our batteries this year, but have not done our new one. I will test that one this Saturday and let you know.

-Hugh

[Chris_Elston]

We do have a Battery Beak Tester as well, we use for a spot check right before the battery goes in the robot.
http://www.andymark.com/product-p/am-0995.htm

But so far it seems a basic CCA rating from our "cheap" automotive tester seems to be a good gauge based on what we see on the field. I am mostly interested in "capacity" of the battery. How long can I run a robot on a fully charged robot before it dies completely.

At 500 CCA - we might get 6-7 minutes
At 800 CCA - we will get 10-15 minutes

That's about some rough numbers we've seen over the years.

[Ken Streeter]

Have you load tested these batteries?

...

1. Do you load test batteries?
2. Did you load test the EnerSys Genesis NP18-12B 2013 KOP and FIRST Choice Batteries?
3. What was your results?
4. Can you confirm if your batteries meet your team spec?

Interesting.

Quick summary to your questions is immediately below. Details farther along

1. Do you load test batteries?
Yes.

2. Did you load test the EnerSys Genesis NP18-12B 2013 KOP and FIRST Choice Batteries?
Yes, but we only have the 1 KOP battery

3. What was your results?
Our new 2013 EnerSys KoP battery gave comparable performance as prior years, with a capacity (in our test setup) of around 12Ah to 14 Ah (I can provide more detail tonight.)

4. Can you confirm if your batteries meet your team spec?
Yes, our 2013 KoP battery seemed to have performance typical of prior year's new batteries.

More details follow...

First off, we started load testing our batteries a few years ago. We used to make use of an automotive load tester like you describe, drawing a high-current load of 150 amps for 15 seconds, similar to what you describe.
Our automotive load tester wasn't very good and didn't have a good CCA read out, however. We were also concerned about the possible battery damage from doing such a harsh test on our batteries.

So, a few years ago we switched to using a West Mountain Radio CBA III. (See http://www.westmountainradio.com/cba.php - the current model is now a CBA IV). With that tester, we now "slow discharge" the battery with a continuous load of 7.0 Amps until the battery voltage drops to 11.0 volts. This is much more benign to the battery and is a test we feel we can do without risking battery damage. When doing this test, the CBA III produces a nice plot, which can be very informative, especially when compared with plots from other batteries and/or the same battery from prior tests. The key metric we use from this test is how many Amp-Hours is sourced from the battery before dropping to 11.0 volts.

However, neither of the above tests really reproduces "FRC Match Conditions." The first case is way too much current too quickly, and the second case is much less current than in a typical match.

We did test our 2013 KoP battery, and I can provide the plot tonight. My recollection is that our test methodology gave a capacity of about 12Ah to 14Ah, which is right in line with what we have seen with brand new batteries in past years. We typically see between 12Ah and 14Ah for a "good battery" with our test methodology. I believe the 18Ah rating from the manufacturer is for an even slower discharge rate (less than 7.0A) to a lower voltage (less than 11.0V), which will successfully extract even more charge from the battery.

From your tests, it seems like the 2013 batteries you have did not do as well with high-load (100A) test, but the one that I tested was fine in the low-load (7A) test that we did.

I would note that in the past with the test that we've been doing, we haven't seen a difference between the Enersys and MK batteries. (We have a mix of both, as FIRST has been putting Enersys ones in the kit, but all our purchases have been MK batteries.)

[Ken Streeter]

... the CBA III produces a nice plot, which can be very informative, especially when compared with plots from other batteries and/or the same battery from prior tests. The key metric we use from this test is how many Amp-Hours is sourced from the battery before dropping to 11.0 volts.

... My recollection is that our test methodology gave a capacity of about 12Ah to 14Ah, which is right in line with what we have seen with brand new batteries in past years.

OK, I found the summary plot I made for our team's batteries as of the 2011 build season (http://www.chiefdelphi.com/media/photos/38359?), which did not include tests of our then-new 2011 batteries. I'll still have to wait until tonight to check to see our 2013 test results, as those results are on a computer which is only at our build location.

This plots shows that "new" batteries were giving a range of about 14Ah to 16Ah with our test methodology (I didn't remember that quite right, apparently).

To understand the plot, we give each battery an identifier (e.g. 2010-A indicating the year of purchase and a letter for the year). The plots are labeled with the battery name and the date of the test.

I would note that as of last season, we were still using 2 of our original 4 batteries from 2005 (our rookie year) as they still held over 12Ah of charge, although we were using these batteries only for practice as they were no longer competition-legal. (I still need to test these batteries to see if they are worth keeping, but I think last season is likely to have finally killed them - 8 years of competition use including many off-season events isn't too bad!)

I'll try to remember to update this thread tonight with a plot of our test results for our 2013 KoP battery.

http://www.chiefdelphi.com/media/img/374/3749bd74786385aeefc31eb2c0887985_l.jpg

[Joe Ross]

I believe the 18Ah rating from the manufacturer is for an even slower discharge rate (less than 7.0A) to a lower voltage (less than 11.0V), which will successfully extract even more charge from the battery.

Battery manufacturers will measure the capacity at 1 amp or less. MK does it at 0.9 amps. There's a nice chart on the data sheet. http://www.mkbattery.com/images/ES17-12.pdf

[theawesome1730]

At 500 CCA - we might get 6-7 minutes
At 800 CCA - we will get 10-15 minutes


That seems awfully low. We have run almost hours at a time during demos at which point we drop to about 9 volts

[Chris_Elston]

That seems awfully low. We have run almost hours at a time during demos at which point we drop to about 9 volts

These times I posted are actual driving practice matches non-stop driving, collecting, shooting, etc... Demo hardly use any juice. We have lasted about an hour as well on one battery sitting still shooting or something easy to do.

[Chris_Elston]

Ken,

You mentioned that 100 Amps might be too harsh and your test at 7 Amps might be too light. But reading the manual on the CBA tester, doesn't it do higher than 7 Amps? I thought I read that it can load up to 40 Amps?

[Ken Streeter]

You mentioned that 100 Amps might be too harsh and your test at 7 Amps might be too light. But reading the manual on the CBA tester, doesn't it do higher than 7 Amps? I thought I read that it can load up to 40 Amps?This is a perfect example of marketing literature stretching the truth a little!

Yes, the CBA III can draw up to 40 Amps of current, but there is also a 100 watt limit. So, up to 40 Amps can be used for a test, but only if the battery being tested is less than 2.5 volts. (40 Amps at 2.5 volts = 100 watts)

For a robot battery (6 cells, 13.2 volts peak, 12.0 volts nominal), 7 Amps is the limit the CBA III will allow. (7 Amps at 13.2 volts = 92.4 watts). There are lots of safety controls in the software to avoid users setting up dangerous tests, and 7 amps is the maximum the software will allow for batteries of the voltage and capacity of our FRC robot batteries.

[Al Skierkiewicz]

Chris,
The specification on both batteries is a peak current (for a short period) over 500 amps for a new fully charged battery. I highly recommend you stop using the CCA tester. It produces significant currents internally for the small batteries. Even at 10 second duration, it may weaken inter-cell wiring or cause damage to the plate and glass mat internal to a cell. If you really need to use the big load, and you can get 500+ amps or better when you first pull the trigger, I think the battery may be OK. Just don't leave it on for 10 seconds.
We have been using the CBAIII for many years and to match manufacturer specs. We set for 7 amps and 8 volts terminal voltage which is the MK graphs cutoff. I tried to test batteries last week only to find that the CBAIII is not supported for Vista but West Mountain will take it back and refurb to correct the problem. If you own one, see the West Mountain site for details and RMA info.
I should add that high resistance connections may also be giving you false readings. Improper crimps or loose hardware on the battery terminals could show up as low CCA numbers. We use a #10 external tooth star washer between the battery and the wire terminal. This will cut through any surface crud on either terminal and prevent the two terminals from sliding around and loosening the hardware.
West Mountain does sell a higher load device if you need to check larger batteries but the extra load is six times more than a CBA.

http://www.westmountainradio.com/cba.php

[Chris_Elston]

Al,

Thank you for your information, you always have good advice for everyone. The CCA tester is only loading the battery at 100 Amps, but it produces the "CCA" reading based on resistance inside is basically what I understand. It's not loading it up to 500 Amps. I wasn't sure if that was what you meant or not, but I just wanted to clarify. When I mentioned computerize, that is the one I was eye-balling, sounds like the CBA is something we should check into.

thank you

[Al Skierkiewicz]

When you call West Mountain, mention my name and FIRST. I have been trying to get them to sponsor teams or discount. I haven't been able to talk to anybody this year yet. I may miss the DuPage hamfest on Sunday. they sometimes have a booth.

[adciv]

1. Do you load test batteries?
2012 was the first year we load tested.

2. Did you load test the EnerSys Genesis NP18-12B 2013 KOP and FIRST Choice Batteries?
Not yet, we plan to.

3. What was your results?
See attached spreadsheet (some batteries listed not tournament legal). We have found some batteries were not manufactured correctly, this may explain the lower CCA you are experiencing. For measuring capacity, our batteries were discharged at a 10A constant current rate. The equipment used was an aviation battery tester we have access to. Looking at the resulting voltage plots, we could identify bad cells and other problems.

When we had a battery with a bad or weak cell, we would quickly recharge the battery in an attempt to recover capacity. I'd have to check what the fast recharge rate used was, but it was at least 10A. By the way, in case anyone thinks this was dangerous, we the facility we had access to was capable of handling any issue which could arise.

4. Can you confirm if your batteries meet your team spec?
We'll let you know when they are tested. At this point our "spec" is no bad cells and >12Ah. We've found new batteries come in at 14-17Ah typically.

[RyanN]

Over the years, Team Fusion has received a few defective EnerSys Genesis batteries. Enough defective ones that we don't even consider using them now.

By defective, I mean the internal resistance is too high to run a motor. They read a good voltage powering the cRIO and other small loads, but as soon as a heavy load is applied, our robot would reset.

Our number's aren't that high though. Maybe 3 or so defective ones? We also had problems with them splitting open on year, while the MK batteries seemed to not split open at all. (2010, side mounted cradle... it was a bad design).

[Al Skierkiewicz]

Guys,
We chose the 7.5 amp/0.4C discharge rate as it gave an fairly accurate calculation of the design AH rating for the battery while using a test that lasts a little over two hours. We experimented with adding a shunt load to increase the discharge current and as expected the AH rating went down nearly identical to published curves from the manufacturer for the higher discharge rate. Remember that the 18 AH rating is generally accepted for discharge rates in 0.1 C range. Discharge rates above 0.4C will give quicker results but will not display anywhere near the rated AH.

[Hugh Meyer]

Chris,

Our new KOP battery tested at 15.3 AmpHours. The Battery Beak Rint was 0.018 ohms and the charge % was 105. This is on the low side of normal. Most of our batteries test in the 14 - 17 range. This was the first test. Sometimes we see a subsequent test of a new battery measure higher.

Attached is the data from the CBA tester. You should get one. They are a great tool!

-Hugh

[Chris_Elston]

Thanks Hugh...so it was LOWER than normal huh? Ok thanks for confirming that.

[Ether]

Does anyone know: What causes the sharp bend in the curve at 11 volts? What is going on inside the battery at that point?

[AhAhPatel]

Over the years, Team Fusion has received a few defective EnerSys Genesis batteries. Enough defective ones that we don't even consider using them now.

By defective, I mean the internal resistance is too high to run a motor. They read a good voltage powering the cRIO and other small loads, but as soon as a heavy load is applied, our robot would reset.

Our number's aren't that high though. Maybe 3 or so defective ones? We also had problems with them splitting open on year, while the MK batteries seemed to not split open at all. (2010, side mounted cradle... it was a bad design).

Are we required to use the batteries in the KOP or are there other options?

[FrankJ]

Are we required to use the batteries in the KOP or are there other options?

You are required to use the battery specified in the rules. So you do not have to use the battery specifically supplied in the KOP, but do have to use one of these two:
A. MK Battery (P/N: ES17-12) or

B. EnerSys (P/N: NP 18-12)

[protoserge]

Be wary of the cheaper "replacement" batteries. While these are pretty much identical to the legal batteries in terms of performance, they are not legal.

I'd still recommend picking a few of the replacement batteries up for driver training and non-competition practice since you won't be abusing your competition batteries and extend their life. They're also about half cost.

[JDL]

I load tested ours. Poor results, dropped from float voltage to 10.9 volts instantly, never even triggered my scopemeter that was set at to trigger at 50 amps. It's back on the charger and will be tested again this weekend.

[Al Skierkiewicz]

JDL,
Instantaneous drops of 2 volts signal a defective cell. Check the archives for one of my previous posts on CBA curves showing several of these types of failures. Although some of the batteries were pretty old at the time, most drops on new batteries are due to mishandling. Often dropped from a distance onto one corner. My intuition tells me that several plates break off the connector plate within the cell resulting in a short across some or all of the cell.

[JDL]

That's what I figured too. The corners looked ok, but the fact they just hand you a battery sans box or anything at kick off doesn't inspire confidence in me that they are well cared for.

[Gregor]

That's what I figured too. The corners looked ok, but the fact they just hand you a battery sans box or anything at kick off doesn't inspire confidence in me that they are well cared fore.

It should have been in your black KoP tote.

[JDL]

It should have been in your black KoP tote.

Oh is that how they did it this year? Some years we've just been handed a loose battery. Still just tossing the thing in with all the other junk ain't great.

[sanddrag]

Before buying new batteries, is there a definitive word on which is better, the genesis or the MK?

[Al Skierkiewicz]

Both batteries are identical in that they are the same size, weight and construction, i.e. AGM. The slight differences in performance specs are due to the different test methods used by the two companies and the specification chosen for amp-hour and discharge. Actually, all batteries in this size will have nearly identical performance and will carry specs in the 18 amp hour range.

[simpsonboy77]

Does anyone know: What causes the sharp bend in the curve at 11 volts? What is going on inside the battery at that point?




I pulled off my chem book from the shelf so I don't give out false information. I am also going to assume a semi-ideal battery, basically I'm going to assume the concentration gradient of everything in the electrolyte will be constant. This is the reason you can overcharge a cell and also over discharge a cell.

If the reaction is moving from left to right side then it is discharging

Positive electrode: PbO2 + H2SO4 + 2H+ + 2e- <> PbSO4  2H2O

Negative electrode: Pb + H2SO4 <> PbSO4 + 2H+ + 2e-   

Overall cell reaction: PbO2 + Pb + 2H2SO4 <> 2PbSO4 + 2H2O

H+ is notation for the hydronium ion, its just hard to type it in here. H30+(aq) e- is for an electron or the hydroxide ion OH-(aq)

If the cell is fully charged then the plates are lead and the electrolyte is sulfuric acid. As the cell becomes more discharged the concentration of H2SO4 will decrease, and the concentration of lead sulfate will increase.

We can apply the Nernst equation.
V(cell) = V(standard cell potential) - RT/(2F)*ln(Q)
V(cell) is the current cell voltage
V(standard cell potential) is how much voltage the cell gives nominally, ic 2 for SLA, 1.6 for Li-ion etc.
R is the gas constant
T is temperature
2 is the number of electrons transferred between anode and cathode per reaction.
F is faraday's constant
Q is very important :)

Q is the reaction quotient, basically its the concentration of sulfuric acid divided by the concentration of lead sulfate. This is simplified from the actual definition because we can ignore other terms as they are not aqueous.

ln(Q) = ln(concentration of sulfuric acid) - ln(concentration of lead sulfate)

What makes batteries really nice is that the sum of the number of lead sulfate and sulfuric acid molecules must be constant. This is assuming the battery isn't venting a whole lot, leaking, or lead sulfate drops to the bottom or significant sulfation on the plates. For the short term these are fine assumptions, and honestly I don't know if I could take all of these into account properly.

Lets say the battery starts with N moles molecules of sulfuric acid and lead sulfate.
N//volume of electrolyte is concentration of sulfuric acid + lead sulfate

ln(Q) = ln(concentration of sulfuric acid) - ln(N/volume of electrolyte - concentration of sulfuric acid)

ln(Q) = ln(concentration of sulfuric acid/(N/volume of electrolyte - concentration of sulfuric acid))
This is going to relate DIRECTLY to how much charge the battery moves.
charge = current times Time. By discharging it reduces the concentration of sulfuric acid in a linear rate, yet the cell voltage is going to be reduced in a logarithmic fashion.

V(cell) = V(standard cell potential) - RT/(2F)*ln(Q)

[Garrett.d.w]

I am assuming everyone has already checked this...but.

In my experience at least, poor performance of a new battery is generally caused by a poor connection with the battery terminals. If you haven't already, clean the terminals with an abrasive and re-seat all of your connections.

[Hugh Meyer]

Ether and Others,

Here is the data from the graph I posted earlier in this thread.

-Hugh

[Ether]

Ether and Others,

Here is the data from the graph I posted earlier in this thread.

-Hugh

Hugh,

Thanks for posting that data.

I played around with it and was able to fit a logarithmic model as Garrett (http://www.chiefdelphi.com/forums/showpost.php?p=1221499&postcount=30) suggested, with a little bit of coercing.

The blue is your data, the red is the model.

It still doesn't like that sharp bend in the curve, so it's not following the model exactly.

@Garrett: if you have suggestions for changing the model let me know and I'll try it to see if it gives a better fit.

[Seth Mallory]

The AH rating of the batteries are listed at a 20 hour rate. The Spec for a 17HR battery is .85 amps for 20 hours and the 18HR is .9 amps. The higher the rate of draw the less you get out of them. I have been using this type of batteries Amateur radio for years. I have ones from 12HR to 55HR and they all are rated at using the 20 hour draw test. My 4 year old 55HR batteries tested at only 45HR this year. Time for 2 new ones.

[JDL]

That's what I figured too. The corners looked ok, but the fact they just hand you a battery sans box or anything at kick off doesn't inspire confidence in me that they are well cared for.

Had the kids leave the new battery on one of our 8 step chargers for the week. Retested it Saturday and it did fine.

[Al Skierkiewicz]

JDL,
Just to be sure, I would run it on last years robot for a while and see if charges and discharges normally.

[simpsonboy77]

Hugh,

Thanks for posting that data.

I played around with it and was able to fit a logarithmic model as Garrett (http://www.chiefdelphi.com/forums/showpost.php?p=1221499&postcount=30) suggested, with a little bit of coercing.

The blue is your data, the red is the model.

It still doesn't like that sharp bend in the curve, so it's not following the model exactly.

@Garrett: if you have suggestions for changing the model let me know and I'll try it to see if it gives a better fit.


I have given it some thought, and I have a few reasons why it wouldn't match correctly, but I'm not sure how to model them. I'll stop by one of my professors (hopefully) by the end of this week and see if they have any input. My current thoughts are:

1. As current is flowing the concentration gradient will not be constant. The higher the current the more skewed the gradient will become.
2. The effective surface area may change. As the battery becomes more and more discharged, each terminal will have lead sulfate on it, which will not react and will hinder the reaction.
3. As the electrolyte becomes more dilute its resistance may change. The plates also will have a changing resistance, especially with sulfation.

If anyone has a battery beak from crosstheroads, or a similar device, could you post what it thinks the internal resistance is of a battery at multiple voltages? I know its alot to ask, but if you could give a sweep from like 13 to 11 at like half a volt steps that would be awesome.

I feel like something is going to have to change which has to do with the rate of reaction. Maybe the plates are not uniform after several cycles.

My team has a battery with a dead cell, I may open it up in a lab and see if I can test some things. I'd discharge the remaining 5 cells and at certain points stop the discharge and measure the resistance of the battery. I would also look into making some of the plates viewable, and either take video or pictures of it at different stages of discharge.

[Al Skierkiewicz]

Garret,
There are several views on the battery internals here and on other areas of the internet. I would not recommend opening the battery. Disposal and cleanup of the electrolyte and the exposure of lead in uncontrolled environments makes me cringe.

[AllenGregoryIV]

When you call West Mountain, mention my name and FIRST. I have been trying to get them to sponsor teams or discount. I haven't been able to talk to anybody this year yet. I may miss the DuPage hamfest on Sunday. they sometimes have a booth.

Before this thread I was able to get West Mountain to donate an older CBAIII to our team since they are no longer selling them and only selling CBAIVs. We have been testing our batteries from last year and have already found some dead ones. I don't know if they will do this for more teams but they were very nice about it when I sent them an email.