Battery System For High End Customer?

[harryn]

Hi, in case it is helpful, my small company builds 48 volt LiFe setups.

We use a very good, commercial (actually marine rated) battery and the suppliers BMS system to make up 48 volt x 130 amp "packs".  You can parallel as many packs as you like and they don't need to be the same age as the control electronics take care of this aspect.  In a multi pack arrangement, they can be networked and if you like, the details down to the cell level can be monitored on a PC.

Each pack, including the BMS are built into a water resistant / impact resistant case.  This is more important than people think because you don't want someone to come in contact with this much potential power and the system will work better with some temperature moderation.

The design is intended for mobile /  RV applications so it is ruggedized for that market, but they can be used in a stationary application as well.  You can imagine though that ruggedized setups designed for mobile use are more expensive to build than something that just sits around, but it isn't terrible.

Technically the batteries can sustain 150 amp continuous discharge, but I try to setup them up to run at C/2 or less typical, 100 amps max. 

IMHO, the greatest strengths are that they will accept fairly high charge rates to 95% + SOC and there is not a need to fully charge them (except occasionally for balancing).  Their greatest weakness ( which is common for LiFe batteries) is they power really falls off at 40 F and if they really go below 35 F or so you can't charge them, so there are protections to prevent this.

I would be happy to supply the packs and act as a sub for your project if interested.

[harryn]

BTW - you don't want to build your own BMS for Li batteries.  It needs to be in control of the pack down to the individual cell level and be capable of balancing across the entire setup or it is useless.

Could you build one that will "work" for a DIY use?  Yes

Would you want to sell that to a customer?  IMHO, absolutely no.

I thought about the same thing and am happy that I decided to avoid this.

[RossW]

BTW - you don't want to build your own BMS for Li batteries.

Just because there is such a difference ... are you referring to

Battery MANAGEMENT system or Battery MONITORING system?

(To be clear, I have a commercial cell BALANCING system, and a home-made battery MONITORING system)
(Attaching picture of a 60-cell monitoring board. Not yet cleaned, so look past the grot on the boards. These also support per-cell temperature monitoring. Everything is on the board except the connector. Yes, that's RS232, yes it's old, but there's so much stuff out there to get RS232 onto a network easily, reliably and cheaply. And yes, it's optically isolated)

[harryn]

Ross = Excellent point.

As you noted, many of the systems marked "BMS" are only "Monitoring", which is sort of like watching your car crash into a wall rather than applying the brakes to keep it from happening.

The BMS I use is doing active management.  There is a BMS board built into each battery, the batteries in each "pack" are networked to a controller for that pack, and then the controllers are networked to a PC if required.  I use this setup because that way each "pack" is independent of the others and has its own control for the pack that deals with charge / discharge control, balancing, etc.. 

At the PC level, it is just monitoring though, not active control so it isn't dependent on windows to work.

Nothing wrong at all with RS-232 control - that is a nice looking board set.

I tried to hire some engineers to build up a board set like you have and write the controls code, but was not able to find a critical size / skill set team to pull it off.  In the end, it has worked out ok for me though because I can use the battery supplier's size and strength to reinforce my efforts, and it enables the customer to find replacement parts if I fall off a cliff.

[RossW]

As you noted, many of the systems marked "BMS" are only "Monitoring", which is sort of like watching your car crash into a wall rather than applying the brakes to keep it from happening.

I've not found the need to take insane precautions with mine, because of a number of (perhaps) special circumstances.
First, cold. I live in an area that rarely gets below -5 deg C outside. But then, my batteries are in my power room which is burried 12 feet underground in a concrete bunker. On the coldest day, it never gets below about 18 deg C in there.

Each bank is fused with suitably rated HRC fuses.

The inverter will call for genset to start if batteries get below about 35% SoC, or volts get down to 48V. If the genset doesn't start, the house will call/page/alarm anyway.

Even if nobody responds to generator failed to start, the inverter has LVD at 46V. The LFP cells are still well within "safe" zone even at that.

The inverter and charge controllers are all set to cease charging safely below Vmax. My cell monitoring will alarm if any cell gets more than 50mV from any other, or above or below a threshold that is well within "safe" limits for my cells.

I already have cell BALANCING (as described earlier).

Because of these factors, I considered the extra overheads and points-of-failure a battery management system (with HVD/LVD/SCP) to be unnecessary and perhaps detrimental to overall system reliability.

The BMS I use is doing active management.  There is a BMS board built into each battery, the batteries in each "pack" are networked to a controller for that pack, and then the controllers are networked to a PC if required.  I use this setup because that way each "pack" is independent of the others and has its own control for the pack that deals with charge / discharge control, balancing, etc.. 

Yep. For many people, this level of autonomy is probably justified (or required!)

Nothing wrong at all with RS-232 control - that is a nice looking board set.
I tried to hire some engineers to build up a board set like you have and write the controls code, but was not able to find a critical size / skill set team to pull it off.

I did it by myself. Designed the circuit, layed the circuit board (ok, the blanks are fabricated by others), assemble them, test them, calibrate them, wrote the code myself...  (and my boards work fine for lead/acid or lithium cells!)

It's all good fun though.

[CDN-VT]

Nice read & understand

VT

[harryn]

Ross - I am very impressed.

[Ron Swanson]

I am impressed as well, but how much of this is applicable to a customer who knows very little, when a service call costs a couple hundred bucks?

Just asking.

[RossW]

how much of this is applicable to a customer who knows very little, when a service call costs a couple hundred bucks?

That's a good question, and one I can't directly answer.
I can however cite one installer over my way who got a 48V set from me for a new install he was doing for a customer.
He installed them and connected them to a small terminal server and a cellular modem, which he also had connected to a Classic, and the main inverter (a well known brand over here in Australia). He was able to monitor the customers system remotely, identified some disturbing behavior from a few of the new AGM cells, was able to shoot all that data to the manufacturer and scheduled a visit with the customer some time later - when he had some replacement cells just in case, and some tests to carry out to eliminate a few possible conditions.

In his case, it probably saved an expensive AGM bank, a lot of customer angst and a whole lot of finger-pointing. Getting on to the problem early (which I might add, did NOT show up on the bank voltage charts - which looked absolutely normal).

How would it help an average end-user? Well, if the installer were the diligent type (like so many of you guys seem to be), the fairly small additional outlay gives a degree of understanding not previously attainable, and could see problems responded to before they become expensive disasters. Sure, FLA is still going to require regular maintenance, but the confidence factor of knowing things are behaving well is hard to put a value on.

[Ron Swanson]

Sure, FLA is still going to require regular maintenance, but the confidence factor of knowing things are behaving well is hard to put a value on.

I think that is the conclusion I am coming to in my mind as well.

1. FLA are cheap and so long as they are easy to get at, several replacements in the space of one replacement cycle of a lithium product might still be money ahead.  If they are not easy to get at that is bad building design and not our problem.

2. With FLA we can see what is going on inside them and make corrections if necessary.  With sealed products it is just interpolation based on terminal voltage and behavior alone.

3. The kind of person that buys a 50+ kWh battery bank ought not to balk at periodic service visits anyway, and what with all the other fun things that can happen, it can pay to be on site every month or two to make sure nothing is falling apart before a midnight phone call comes in.

Thoughts?

[RossW]

Not everyones perspectives, motives, expectations and environments will be the same.
Mine is probably a significant distance from "the norm" (whatever that is!)

1. FLA are cheap and so long as they are easy to get at, several replacements in the space of one replacement cycle of a lithium product might still be money ahead.  If they are not easy to get at that is bad building design and not our problem.

I hated the bulk, the maintenance, the acid mist, the constant messing, checking, tweeking, etc of my small FLA cells. (8 x T105 batteries in series). I hated how "soft" they were, couldn't afford or justify a huge bank that would have been stiffer (but that I wouldn't have had enough charging capacity to keep healthy anyway), I resented that >15% of my expensive fuel for the genset was "wasted" by the batteries intrinsic cycle inefficiency, etc.

2. With FLA we can see what is going on inside them and make corrections if necessary.  With sealed products it is just interpolation based on terminal voltage and behavior alone.

Well, yes, you can see what's going on inside if you're constantly peering inside them, which I didn't want to be (and couldn't afford to pay someone else to do, even if there WAS someone within managable distance)

3. The kind of person that buys a 50+ kWh battery bank ought not to balk at periodic service visits anyway

Well, I actually take offense at the implications of that. "He's got enough money to get the good stuff, we can screw him on constant hand-holding, extra services, etc". Sure, that might apply to some, but for others of us - it was a very hard decision, a huge investment in stuff that I really could have used the money on other things, but that ultimately the fact that they DON'T require extra servicing, checking, poking and prodding for a "mere few hundred bucks a month", became one of the final straws.

I have a "mere" 48kWh storage - so perhaps don't make it into your category anyway  but the things I had to forgo in order to do it are non-trivial.  I'd be surprised if I'm the only one in this situation.

with all the other fun things that can happen, it can pay to be on site every month or two to make sure nothing is falling apart before a midnight phone call comes in.

Thoughts?

I spent a lot of time "gaining confidence" with my system. Checking the graphs. Checking cell temperatures, currents, making sure each cell WAS actually at the right voltage, and that the delta-V across each cell in the pack really was as low as my system said. As I gained confidence, I checked less. I made my system monitor and spit me out daily health checks. I reduced my burden further by making another system check it got the daily messages (which I then increased to every 4 hours) and calls me if it doesn't get the status message OR if anything in the system wasn't right. The monitoring system will call me within a few minutes of any cell going outside limits. This stuff can all be easily automated, and once automated gives you basically continuous and continuing information and confidence. Long-term graphing of data will (I hope) indicate any gradual change in behavior - voltages creeping up or down, charging temperatures not tracking ambient as well as they used to, an individual cell getting hotter - all this stuff when viewed over a year make changes stand out. And that, to my way of thinking, is where we probably all should be aiming to be. There are so many small, low-(electrical)-power computers with amazing grunt, that this sort of thing will (I believe) be the norm in the near future.

Once that's in place, if the cells are FLA, AGM, LFP, NiFe, NiCd or something else probably won't be so much of an issue, as simply using a technology that's "appropriate for the site". I love my LFP because they are so stiff, and for the daily usable energy - so small and so light. On a houseboat, RV or caravan, that's probably a major consideration. Corrosive, spillable electrolytes may be ok in some places and not so in others. Lead is being frowned on in lots of parts of the world. While it is still something you can sell as scrap, I forsee a day not too far away where you won't be able to transport them without a hazardous materials permit (already happening in some places), where you won't be able to sell them except to a licensed hazardous materials recycling facility (happening now), and you won't get paid for the lead - you will have to pay for them to take it and dispose of it. When will that happen? Will it be before your next set of batteries die? If so, what extra cost to dispose of them?

(The same arguments probably apply in varying degrees to most battery technology available now)

[krementz]

We may also be able to get good deals on Deka Unigy.  Never seen them in the field yet.

Thoughts?

I have had a 52 V 1804 A/hr string of Deka Unigys for about 2 years. So far, no problems.

[Ron Swanson]

Well, yes, you can see what's going on inside if you're constantly peering inside them, which I didn't want to be (and couldn't afford to pay someone else to do, even if there WAS someone within managable distance)

3. The kind of person that buys a 50+ kWh battery bank ought not to balk at periodic service visits anyway

Well, I actually take offense at the implications of that. "He's got enough money to get the good stuff, we can screw him on constant hand-holding, extra services, etc". Sure, that might apply to some, but for others of us - it was a very hard decision, a huge investment in stuff that I really could have used the money on other things, but that ultimately the fact that they DON'T require extra servicing, checking, poking and prodding for a "mere few hundred bucks a month", became one of the final straws.

I have a "mere" 48kWh storage - so perhaps don't make it into your category anyway  but the things I had to forgo in order to do it are non-trivial.  I'd be surprised if I'm the only one in this situation.

You obviously know "how to do it" and I am impressed by your success.  And your chemistry may make it onto the major markets if it is that good.  I too don't like all the things you listed above, that is why I asked the original questions in post #1.

But with all respect, I wonder about the applicability of your experience to a relationship between a service company such as ours and a customer who just wants power for his house which he chose to plop in the middle of nowhere.

Our concern is for our own profitability, so we will still be around next year to help customers, of course, but we have decided that the #1 way to do that is to do what makes sense for the customer. 

So this question is really, what gets the customer the lowest cost per kWh delivered, when they do not want to put their own hands on the system.  These are the same people that do not do their own gardening or mow their own lawn.  They do their work such as running some large company or creating art or music or something that earns them more than we ever hope to, but such is life.

Is this all making sense?  Do you agree or disagree?  Can you give some part numbers, rough dollar figures, and hours required to set up a battery/inverter/generator setup like you have?

[RossW]

So this question is really, what gets the customer the lowest cost per kWh delivered, when they do not want to put their own hands on the system.

I still think there's more in that equation than merely "lowest cost per kWh", but yes, I see what you're trying to do.

These are the same people that do not do their own gardening or mow their own lawn.

There are lots like it. I'd like to be one of them I guess

Is this all making sense?  Do you agree or disagree?

It has, right from the get-go. I agree with most of it - at least in principle. As always, the devil is in the detail, and there isn't going to be any one solution that works for all cases.

Can you give some part numbers, rough dollar figures, and hours required to set up a battery/inverter/generator setup like you have?

Not off the top of my head. Mine has been an evolving system for 15 years. Adapting to my changing needs; to changes in available technology; to prices of fuel, PV, storage, anticipated disposal costs and my own funds.
When we started building here, offgrid was never the intention. It wasn't until we were ready to "break ground" that the changes in local regulation meant that the "right of access" for power companies was no longer a given and we found ourselves unable to get the grid - so it was either completely abandon the entire project, or suck it up and make do.
The initial build (err, clarify that - batteries/generator/inverter) was about $20K (AUD). The genset was fairly expensive because it was a custom build co-gen (actually tri-gen). The batteries were what was available at the time (used AGM).
At $10/W (at the time) PV was basically non-existant. After the price of our sole energy input (propane) increased significantly, and the price of PV fell to $6/W, it actually made sense to scratch together the $8K it cost me to import some panels and supplement our energy with PV. Paid for itself in 12 months in saved propane. More batteries, more PV, better inverter, better batteries, even more PV, extra inverter - it's been growing and adapting.

[harryn]

Ron, you asked about some price estimates.

I sent a PM to you based on:
- 10 - 12 kW continuous inverter name plate capacity for a USA 240 vac split phase setup.  (there is surge capacity as well)
- 48 kW-hr usable battery capacity (not nameplate, actual usable capacity)
- Packaged the way we do it for ruggedized / mobile use which is also fairly weather and water resistant.

Obviously the ruggedized approach we use makes the system as a whole more expensive, so it might not be viable for stationary applications - not sure.

Interestingly, in this capacity range, the price difference between LiFe and AGM was surprisingly small.

Harry