96 or 144v instead of 48v

[sammy_sparrow]

Mucho better IMO. Even 3 hours @ 6000 watts is a beast of a battery 48 volts @ 750 AH. About 1900 pounds in Pb or 900 pounds in a 48 volt @ 600 AH LFP

PB batteries have lasted 7 to 10 years and have done it for the past 100 years. No Chi-Com LFP is known to last more than 3 or 4 years yet. Lots of promises and claims, but no history of doing it. Very poor track record so far.

I could not sleep knowing that. Can you? Only way I would do it is if the customer signed a Release.

yeah, food for thought....very very serious thought indeed.

[sammy_sparrow]

At 6000 watts, 48 volts will work but you are near the top of 48 volt power range. That is the key is maximum Power, and the current required to push it at a specified voltage. At 48 volts with 6000 watt load you are talking about 125 amps of current and a minimum size FLA battery of 900 AH. So you are already locked into 2400 pound $8000 battery that will only give you 4 hours run time. That battery will need replace din a few short years.

SK, are we talking about the same thing? I am talking about keeping the inverter at 48v input - ie, from a 48v battery bank to the inverter. a 900AH Pb battery at 48v feeding the inverter would let me power the 6000 watts with ease. Heck, right now i have a 48v input inverter with 200AH of AGM on it at home and pull 4500 watts off it without issues.

Are you talking about 6000 watts *AT* 48v?

[sammy_sparrow]

Seems like one or more UPS's plus your own genset plus the building genset plus the grid should give you four layers of redundancy.
Would that be cheaper than the big battery you were thinking about?

It would - but hard to manage. Not sure how i would manage such a setup - it would be lots of manual plugging and unplugging. Not sure it works.

[Sunking]

i like the - 3 hours UPTIME and genset thereafter routine. I think that works.

Mucho better IMO. Even 3 hours @ 6000 watts is a beast of a battery 48 volts @ 750 AH. About 1900 pounds in Pb or 900 pounds in a 48 volt @ 600 AH LFP

So which battery should i look at ? Pb or LFP?

Only way I could sleep at night is with Pb. I know with Pb will cost me half of LFP, save even more by not having to purchase all the auxiliary equipment required for LFP, and proven track record that will in all likelihood will last twice as long. Lots of PB batteries have lasted 7 to 10 years and have done it for the past 100 years. No Chi-Com LFP is known to last more than 3 or 4 years yet. Lots of promises and claims, but no history of doing it. Very poor track record so far.

I could not sleep knowing that. Can you? Only way I would do it is if the customer signed a Release.

[DanKegel]

Seems like one or more UPS's plus your own genset plus the building genset plus the grid should give you four layers of redundancy.
Would that be cheaper than the big battery you were thinking about?

[Sunking]

Your advice is sound. Thank you for your input!

So bottom line here is stick with 48v.

At 6000 watts, 48 volts will work but you are near the top of 48 volt power range. That is the key is maximum Power, and the current required to push it at a specified voltage. At 48 volts with 6000 watt load you are talking about 125 amps of current and a minimum size FLA battery of 900 AH. So you are already locked into 2400 pound $8000 battery that will only give you 4 hours run time. That battery will need replace din a few short years.

I will be striving to build out 8 hours of UP time from the batteries. I would like an Exeltech MX System to get N+1 on the inverter.

Sammy do you understand what you are asking for? 6000 watts x 8 hours = 48 Kwh. At 50% DOD, a line you do not want to cross with Pb you are talking about a 96 Kwh battery. Do the math on that. A Pb battery weight 60 lbs per Kwh and cost $200 to $250 per Kwh. 3 years replace. Double what I quoted above. You are talking about a 5000 pound $20K battery every few years.

I've been looking at commercial LFP but very expensive. Is it practical to consider CALB's and put these in place ourselves? Exeltech can do the inverting, power comes from the CALB's - just need to figure the charging. From a long time lurking here i wont talk about "balancing", but bottom seems easy enough. does that sound OK?

For a Commercial System using LFP, you want Top Balance because it is made for consumers and morons. It takes an expert to use a BB LFP. Pretty much takes the same equipment either way.

As for CALB my friend you are gambling. Yes I know there are claims they will do 2000 cycles to 80% DOD. Those claims have been there since Thundersky began manufacturing. They went bankrupt because no battery they made ever seen 500 cycles. Here we are 9 years later, and on the 4th revision. None have ever seen more than 500 cycles yet. You would be betting the Farm the latest revision will do what they claim. Do you really want to trust a bunch of Red Chi-Coms who have a terrible track record and known to be liars? Be careful with Chi-Coms. None have a proven track record. CALB, Sinopoly/Winston all are regurgitated Thundersky batteries.

Here in the USA CALBS are used by DIY EV builders. None have ever been able to file a Warranty Claim. Be careful. With LFP you can use a smaller AH battery. Run them between 10 and 90% DOD. Never fully charge them, they will last twice as long if you stay between 10/90. To replace a Pb with LFP for every 100 AH of Pb takes about 70 AH of LFP to replace it. Yes smaller and lighter, but more expensive.

Keep this thought in mind. You can get a top of the line Pb battery for around $220 Kwh. Assuming professional care an dTLC you will have a 7-year battery with full warranty from proven technology from reputable companies like Rolls, Crown, and Exide to name a few. A Chi-Com LFP will run you $450 to $500 Kwh and one can only guess how long they might last. Its a gamble. So be careful, your reputation is on the line. If done in the USA, and the batteries fail prematurely, the replacement cost comes out of your pocket. You would be held accountable. Chi-Coms don't give a damn if it bankrupts you, they are untouchable.

Good Luck to you.

[Mike90250]

Aside from the UL / Code defining <50V. Much of the "efficient" gear, is targeting the 50V ballpark. Higher than that, it's not geared for off grid, and efficiency is traded off for cost savings. So controllers and inverters may only be 80% instead of +90% efficient. Data centers have huge power budgets, unlike off-grid installs or Grid-Tie, where efficiency counts a lot.

[inetdog]

still yawning here - off on a trip out into the bundus (ie, bush out back of nowhere ) today so an early start. First cup of coffee still going in.

Thinking a bit more about it, why even have a fuel trailer? Why not just keep 44 gallon drums at home full and ready to go on a normal trailer - saves cost on having a specialist fuel trailer - more money for other nice stuff in the teleport.

i like the - 3 hours UPTIME and genset thereafter routine. I think that works.

So which battery should i look at ? Pb or LFP?

Only looking at cost for a given amount of energy storage, Pb hands down.
If you have special needs like low weight for a given energy or the ability to sit at a partial charge state without damage, definitely Lithium. LFP is just one Lithium chemistry option, but is the safest and most stable that is widely available and has a long track record. Not the highest power density or the lowest cost per watt-hour though.
Another factor that favors Li is the need to recharge fully over a short time period (several hours, for example). You can go with AGM for that, and pay more than for equivalent capacity FLA or you can go to Li.

[sammy_sparrow]

still yawning here - off on a trip out into the bundus (ie, bush out back of nowhere ) today so an early start. First cup of coffee still going in.

Thinking a bit more about it, why even have a fuel trailer? Why not just keep 44 gallon drums at home full and ready to go on a normal trailer - saves cost on having a specialist fuel trailer - more money for other nice stuff in the teleport.

i like the - 3 hours UPTIME and genset thereafter routine. I think that works.

So which battery should i look at ? Pb or LFP?

[sammy_sparrow]

Lot's of ways to do this, just some ways are a lot better than others. Solar would never be a consideration. You cannot use RE with Mission Critical applications unless you overkill the crap out of everything and NO DAMN BUDGET to get in the way. That is your Rock and Hard Place. You cannot have your cake and eat it too, or Champagne Taste on a Beer Budget.

Good Luck to you. SK

And therein lays the truth!!

Your advice is sound. Thank you for your input!

So bottom line here is stick with 48v. I will be striving to build out 8 hours of UP time from the batteries. I would like an Exeltech MX System to get N+1 on the inverter.

I think the only sensible thing to do re the genset is get our own and manage it properly. Still cant store much extra fuel so will have to think outside the box on that one - something along the lines of a fuel trailer with 1000 liters onboard and keep it at home and on standby ready to go service the genset when the situation arises.

I've been looking at commercial LFP but very expensive. Is it practical to consider CALB's and put these in place ourselves? Exeltech can do the inverting, power comes from the CALB's - just need to figure the charging. From a long time lurking here i wont talk about "balancing", but bottom seems easy enough. does that sound OK?

[Sunking]

Ok so in my particular situation this is not telecom equipment. Its for a communications setup but its not telecom. All the equipment is mains powered.

Telecom uses basically 3 voltages. 24, 48, and 480/270 3-phase. A modern Data-Center will always have both 48 and Mains for the servers. It is a moot point.

In all of this there is the theory and the real world. I know the theory says do a power budget - which I have - hence the 6Kw figure. The important question of how long do we want to power the equipment ? - there is such much in that statement ! In a perfect world, i'd like to run for a week on battery so i have lots of time available to find a solution to a genset that is down or a fuel shortage happening and taking days to find fuel. In the real world we have budgets to work within and so that determines how much back up time we can buy.

I agree and until you determine how many watt hours you need in a day, al this is just academic, not real. Until you get real numbers. I cannot help.

We can argue till the cows come home on the "correct way" to do it but the real world says, we buy as much battery as we can afford, whatever that comes too, and then learn to deal with the shortfall. We will never have enough battery to keep the load UP as long as we would ideally like. We are in the middle of a third world country. The local grid has many "challenges". The genset within the building we are in is not as well managed as it should be and is not 99% reliable. We have load shedding happening every day. Sometime we get hit twice in a day with load shedding. We get fuel shortages at times. There is diesel stored on site but that is only so big - reality is they cant store more. It is what it is. We have to deal with it. So how much battery do we put in? Good question - just how many watt hours do the batteries need to be able to deliver under all these circumstances? So i hope you begin to see its not quite as easy as you are making it out to be - maybe in the USA or Europe it is, but it surely is not here under these conditions.

Your battery capacity depends on how often you want to replace them. Cycle life is directly related to how deep you discharge your batteries. A good AGM may get 1000 cycle if you limit discharge 20% DOD. If you discharge to 50% the maximum recommended DOD before recharge you are down to 250 cycles. You are going to pay up one way or the other. No free rides in physics. Everything must be bought and paid for in advance. No exceptions.

Other real world restraints is the amount of weight the building owner will allow us to add. We are on the forth floor of the building. I cant add 4 or 5 tons of battery. You and I can argue till the cows come home about how stupid this is - the building can surely hold a LOT more weight then that - following the logic of the owner i should never have 50 or 60 people on the forth floor? Crazy ! But again this is not the western world we are dealing with here and a college education can be rare here, again it is what it is, nuff said !! - 2V 6000 Amp/Hr batteries is simply ridiculous under these conditions. With LFP the weight and space savings are real. It may not be an issue to most setups but i have some unique issues to deal with - real world issues - not text book lecturing. LFP really seems to be an advantage here regards weight and space to house the setup. Also i am in a situation where i have a section of the forth floor. There are other users on the forth floor - i dont have it all to myself (one day i'll get it all but for now its shared). I cant run FLA because of the issues of gassing. Have to keep it safe. Also do not want to deal with any off gassing from the batteries near my expensive equipment. So if i do use Pb it will have to be AGM or GEL , and of course it wont be GEL, so AGM it is if we go with Pb. Yes AGM off gasses as well but its tiny tiny compared to FLA and so acceptable if we have to go that way. Again LFP have the advantage here.

If charge voltages are set correctly and current limited, FLA do not gas to any significant extent. Only when you go above gassing voltages. do FLA gas, and that is around 2.4 to 2.5 vpc.

Lithium will provide you some advantages, and would be my choice over AGM. But there is a catch to LFP. There is not much in the way of equipment that is designed to work with it in your application. The advantages of LFP are higher Specific Energy Density. About 1/2 the weight for a give capacity, at 1/3 the volume. They can also be operated in Partial State of Charge unlike any Pb battery which must remain fully charged at all times. In fact to get the most life out of LFP is to never charge above 90% or allow to go below 10%.

With lead acid how fast a charge rate can i hit them with while getting the balance between serving the mission and getting some sort of economical life out of the battery pack? The mission is all about getting a return on investment for the service provided - cant simply say, forget the economics of how long the batteries last if you are charging them really hard to get back to full charge before the next "lights out" event. If we have a fuel shortage and run out of reserve fuel then it will be batteries only - power is generally off for 8 hours once per day. So we have 16 hours till next" lights out". But sometimes - and no figuring when it will happen....... it just happens now and then, we get a second round of "lights out" in a day. So sometimes we only get 8 hours of grid in 24 hours. So i do need to be able to charge up to full charge within 8 hours. The bigger the Pb battery the more current we can put into them, but there is a point where we have to consider are we doing damage to the batteries with a really high charge current. Again, LFP seems to have the advantage here.

Charge rates vary. There are three basic Pb batteries. SLI aka Starting Lighting & Ignitions which I will not address other than to say they can be charged very fast, but has no meaningful cycle life, so that is a dead end.

A True Deep Cycle battery has very heavy thick plates that allow a long cycle life. As a consequence means th eInternal Resistance is higher than we would like and limits charge and discharge rates to C/8.

Then there are hybrids which try to be both SLI and Deep Cycle. The trade-off is you do not get the cycle life of Deep Cycle, or the very high C-Rates of a SLI battery . Hybrids as a rule of thumb are C/4. LFP depending on what manufacture and type can be as high as 10C.

So here is the problem How the Frack are you going to get a Pb battery completely recharged in a short time? You need at least 10 solid hours or unlimited and uninterrupted power. Solar cannot do that, only a Genny or commercial power can do that.

Here are the useful facts of FLA vs AGM.

FLA has a minimum of 2 times the cycle life of AGM, and AGM cost 2 times more than FLA. So the AGM solution is 400% higher in cost, about the same as LFP. So you are in between a Rock and Hard Place. You cannot likely afford AGM and refuse to use FLA because of misconception of gassing. You cannot have it both ways.

Here in the USA here is how we build a Mission Critical Data Center, Telecom Offices, and Hospitals with 99.5% availability.

Dual AC service (two separate services, redundant from two different Sub Stations and Power Plants. ), 3-phase, 4-wire 480/270. Dual Diesel Redundant Generators, Dual Redundant Parallel Dual Conversion UPS. 48 Volt DC plant with direct feed from utility or Genny. 48 volt battery plant is sized for 4 to 6 hours to allow time for mechanics to get generator on line. Diesel genny's have two tanks; Day Tank and a weeks worth of fuel in storage tanks. fUPS batteries are only sized for 15 to 30 minutes using AGM. Special; AGm batteries are used to deliver 4C discharge current, and charged at 1C. Trade off is the AGM batteries only have about 100 to 150 cycles in them so they get replaced every 5 years whether they are used or not as 5 year sis beyond calendar life. Bottom line there is no budget, it is what it is to get 99.5% reliability. A data center generates over a million dollars per hour revenue. A few extra million in equipment is no problem and a drop in the bucket. No Data Center in the USA uses Solar or renewable energy except Google, and that is for show only, does not actually contribute much of anything. Strictly PR as Google is flushed with more cash than they know what to do with.

If this were me, I would size the batteries to carry the load only for a few hours like a battery plant. I would use a battery that can handle a C/4 charge and discharge current. So here is how it would work. If commercial power fails the batteries are already on-line because I would use a Dual Conversion UPS meaning AC>DC>AC or Rectifiers > Batteries > Inverter. When batteries reach 50% DOD, genny starts up, picks up the load, and recharges the batteries in 3 hours and shuts off. Repeat. If the outage is shorter than 2 hours, and comes back on, commercial power recharges the batteries in 1 to 3 hours and ready to go again.

Lot's of ways to do this, just some ways are a lot better than others. Solar would never be a consideration. You cannot use RE with Mission Critical applications unless you overkill the crap out of everything and NO DAMN BUDGET to get in the way. That is your Rock and Hard Place. You cannot have your cake and eat it too, or Champagne Taste on a Beer Budget.

Good Luck to you.

SK

[sammy_sparrow]

Jason I can help you, as I design Data and Communications centers.

Hi Sun King. Great, look forward to your advice. Not a lot of experience this side with power systems, no. Have done minor setups with inverters/batteries but only 24v with 200 amp/hr of battery. Small stuff. Good experience to see how it performed though. Still it gives one an idea of how little is actually in those batteries.

Ok so in my particular situation this is not telecom equipment. Its for a communications setup but its not telecom. All the equipment is mains powered. Totally agree with you that this leads to double conversions and therefore waste, but that's how it is. None of the equipment is 48v. While there are some benefits of using 48v power straight from a DC-DC converter that is connected directly to the battery pack you cant argue with the simplicity of using mains powered equipment because of the versatility of sourcing equipment and parts from just about anywhere. You have a failure on a 48v BUS system or DC load distribution system - good luck finding parts for that out here. Even finding an electrician who knows what he is doing with DC systems may be difficult.

In all of this there is the theory and the real world. I know the theory says do a power budget - which I have - hence the 6Kw figure. The important question of how long do we want to power the equipment ? - there is such much in that statement ! In a perfect world, i'd like to run for a week on battery so i have lots of time available to find a solution to a genset that is down or a fuel shortage happening and taking days to find fuel. In the real world we have budgets to work within and so that determines how much back up time we can buy.

We can argue till the cows come home on the "correct way" to do it but the real world says, we buy as much battery as we can afford, whatever that comes too, and then learn to deal with the shortfall. We will never have enough battery to keep the load UP as long as we would ideally like. We are in the middle of a third world country. The local grid has many "challenges". The genset within the building we are in is not as well managed as it should be and is not 99% reliable. We have load shedding happening every day. Sometime we get hit twice in a day with load shedding. We get fuel shortages at times. There is diesel stored on site but that is only so big - reality is they cant store more. It is what it is. We have to deal with it. So how much battery do we put in? Good question - just how many watt hours do the batteries need to be able to deliver under all these circumstances? So i hope you begin to see its not quite as easy as you are making it out to be - maybe in the USA or Europe it is, but it surely is not here under these conditions.

Other real world restraints is the amount of weight the building owner will allow us to add. We are on the forth floor of the building. I cant add 4 or 5 tons of battery. You and I can argue till the cows come home about how stupid this is - the building can surely hold a LOT more weight then that - following the logic of the owner i should never have 50 or 60 people on the forth floor? Crazy ! But again this is not the western world we are dealing with here and a college education can be rare here, again it is what it is, nuff said !! - 2V 6000 Amp/Hr batteries is simply ridiculous under these conditions. With LFP the weight and space savings are real. It may not be an issue to most setups but i have some unique issues to deal with - real world issues - not text book lecturing. LFP really seems to be an advantage here regards weight and space to house the setup. Also i am in a situation where i have a section of the forth floor. There are other users on the forth floor - i dont have it all to myself (one day i'll get it all but for now its shared). I cant run FLA because of the issues of gassing. Have to keep it safe. Also do not want to deal with any off gassing from the batteries near my expensive equipment. So if i do use Pb it will have to be AGM or GEL , and of course it wont be GEL, so AGM it is if we go with Pb. Yes AGM off gasses as well but its tiny tiny compared to FLA and so acceptable if we have to go that way. Again LFP have the advantage here.

With lead acid how fast a charge rate can i hit them with while getting the balance between serving the mission and getting some sort of economical life out of the battery pack? The mission is all about getting a return on investment for the service provided - cant simply say, forget the economics of how long the batteries last if you are charging them really hard to get back to full charge before the next "lights out" event. If we have a fuel shortage and run out of reserve fuel then it will be batteries only - power is generally off for 8 hours once per day. So we have 16 hours till next" lights out". But sometimes - and no figuring when it will happen....... it just happens now and then, we get a second round of "lights out" in a day. So sometimes we only get 8 hours of grid in 24 hours. So i do need to be able to charge up to full charge within 8 hours. The bigger the Pb battery the more current we can put into them, but there is a point where we have to consider are we doing damage to the batteries with a really high charge current. Again, LFP seems to have the advantage here.

So given a 6000 Amp/hr battery is not possible (weight/cost) we come back to the reality of 2000 amp/hr 2V cells. Even then, 24 of those weigh abut 3-4 tons - so getting into trouble with weight again. With Pb, if i have 2000 amp/hrs i can still not use more then 1000 amp/hrs if possible - go lower and lose battery cycles fast. Ideally we should be sizing to use no more then 20-30% of Pb bank size - no matter if FLA or AGM or whatever. Again LFP seems to have the advantage here - i can use 50% easily - in fact the batteries will supposedly gain many cycles of use if i can keep it to SOC of between 80% and 30%. That gives me 50% of the capacity i can safely use. I f i need to push down lower - 90% DOD, i can if i really need to and the penalty is not that great. Go below 50% with Pb and the loss in cycles is real. Again LFP seems to have the advantage in this respect.

Sunking, you have told me what i stated about AGM's in a data center is "HOG WASH" because in every data center YOU have built you have used FLA. Well, we are not talking about a data center YOU have built - we are talking about a data center (for want of a better description) *I* am building with some really unique issues to deal with. Its not HOG WASH at all about FLA being a no go in *MY* situation.

But hey, i am all ears. I am no expert in this, hence asking questions. I will be the builder of the system but not the designer or specifier.

My original question was about higher voltages so as to avoid parallel strings. Parallel strings dont seem to be such a problem with LFP but are a BIG problem on Pb - right? So given big weights on huge 2V cells are not going to work for me and also that i need to avoid FLA if possible because of the situation of where the installation is, does 96v or 144v using AGMS make sense? What ABOUT if LFP *IS* the right way to go here ( and it seems to me LFP provides answers to some issues i am facing), is it best to stick to 48v and use several parallel strings?

[sammy_sparrow]

For the business end - I'd make sure that whatever you do is compliant with the equivalent of OSHA laws/regulations.
And I'd plan on paying an engineer to design the installation.

Agree with that. I am in the middle of the dark continent so things are a bit wild and woolly at the best of times - changing slowly but we don't have western standards by any stretch.

I am looking at different ways to put this together from prepackaged LFP battery systems to buy in LFP cells and assemble locally. Whatever i do it will be with an experienced engineer to guide and direct us. What i have been doing is reading everything i can so i can make informed decisions on which way i want to head.

The reason i started this thread was to try to find out why is it that 96 or 144v is not more popular and/or used. Starting to get a feel for why now.

[jflorey2]

What are the practical issues in going to say, a 144v inverter system?

Lack of standard switches/chargers/monitors would be a big one. Different safety standards would be a second.

Any issues charging a higher voltage bank from solar array?

You'd be looking at a VERY limited set of charge controllers. Only one I can think of off the top of my head is the Midnite Classic 250KS.

Why are higher voltages not more common in general for off grid and back up power systems?

Because:

1) Traditionally batteries came in 6 or 12 volt chunks, so multiples of those were easy to design to
1a) Traditionally solar panels came in 12 volt chunks as well (18V actually) - no longer true
2) 48V is the highest voltage you can use before additional safety requirements kick in
3) 48V is high enough that reasonable wiring (250 amp breakers, 4/0 wire) gives you enough power for a home (10-12kW)

In the mid-future, as 400 volt DC building power becomes more common, expect to see battery and inverter systems set up to support this higher voltage. However that's still a ways off.

[Sunking]

Jason I can help you, as I design Data and Communications centers. I can tell you do not have much experience because about everything you are saying is not accurate. We use a lot of Exeltech at remote cell sites on solar.

First thing to clear up is what Voltage to select. Most telecom equipment uses either 24 or 48 volts. Servers use AC power which is pretty stupid on the manufacturers part. It means a UPS or another egg in the basket to break down. Anyway for the UPS what battery voltage selected has to do with the load demand. You say 6 Kw which is really small potatoes. At a 6 Kw load at 48 volts is 125 amps and well within reasonable current capacity.

So for now forget about voltage because you have not answered the important question, how many Watt Hours do the batteries need to be able to deliver? You have only told us the demand of 6 Kw which is meaningless. That determines the battery size.

As for parallel batteries, you have not determined if it is needed or not. You can obtain up to 2 volt 6000 AH cells. At 48 volts has a capacity of 288 Kwh. Enough to run a 6 Kw load 48 hours. What this tells you is you have a wrong assumption assuming 48 volts is out because it needs parallel batteries. That is completely false.

Next you said AGm must be used because it is a data center. Hog Wash. Every data center I have built uses Flooded Batteries on the Telecom, and AGM on the UPS. AGM is used on UPS for two reeasons.

1. They are small batteries designed to only carry the load for 15 to 60 minutes. just long enough to get the generator running and transfer load.
2. They small batteries are installed in cabinets and replaced every 5 years.

Flooded batteries in Telecom are massive and last 20 to 30 years.

So tap the brakes a bit because you are about to make some really big mistakes. Do quite a bit more research, ask questions, and do not rush into anything. As of right now all your assumptions are incorrect.