Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
48V LTO Off Grid system
#41
(11-11-2018, 07:17 AM)mblowes Wrote: Korishan.. I'm more incremental.. WVO is a step a bit far for now.

A lot of people in the tropics, where the boat is now, run their gensets over night to keep the  air con on with temps 28-30 deg C.  Problem is that most gensets are oversized for that so there is a lot of wasted energy.  Not to mention the noise.

 Mr mblowes: 
 In order to give better suggestions, Please give more info on :
1) size of ac unit
2) size of genset with hp rating
3) @ cabin sq footage
4)@ ocean water temp
5) cabin water heater tank size
6) battery bank size and volts and type of battery you have now
7) @ sq ft of space for solar cells
8) Most important , what is your skill set background  (in order to make it easy as possible and cost effective)
I mainly work with the Kiss principle (Keep it simple stupid ). When want I learn a trade , I go to the extent becoming a
professional at it. I have been learning for 60 years, Machine shop, Air conditioning, Pluming , Concrete work, Computer building , But spelling and grammar still allude me - tks for spell check
The way I learn is to ask a lot of questions , most people that know will help
Reply
#42
I now have a 36kWh LTO setup split into 2kWh packs, each with a separate 100A breaker.

I am currently sat in a 20ft container in a forest with 6 of the packs (12kWh) powering my computer with 1.3kW of solar laid flat on a hill and a 700W wind turbine sounding as if it is on the runway throttling up to fly to miami peaking just over 1kW in a storm. The rest of the packs are at home providing a base power charged from 7kW of solar.

My pack configuration is 11s2p. The reason for this is that it closely matches 12V lead-acid battery voltage range and I can use the packs with anything that would run off lead-acid or is typically 24V or 48V rated. The 2p was basically to limit the weight of the packs so that I could still carry them as each cell is 1.3kg so 22 cell with wiring and case is just under 30kg. Larger packs then become immobile, have issues with flexing stresses and start to get scary with the short circuit currents.

The one benefit I have found with using aluminium as busbars is that they don't "stick" if they are shorted, i.e. copper bars tend to weld to and grab what they connect to if the current is not high enough to vapourise the material and that's when it gets into a run away moment. With aluminium you get a lot more sparks but less risk, aluminium is cheaper than copper and the studs on the cells are aluminium anyway. Keep it dry and you have zero issues.

Using smaller packs would make for the ability to build and add capacity over time and the smaller breakers per pack for fault conditions. Balancing is a separate issue / story.

Charging - solar depends on the Vmp of the panels, your cell configuration and how much power you are happy to loose. I have had my pannels just wired direct to the battery via a diode on a few occasions and the PWM controllers I use at the moment (for the 12kWh here) are effectively just a glorified switch as they don't buck or boost the voltage. My Vmp is around 10% above the maximum 48V pack charge level (55V) so you waste more solar when the battery is flat and the V pack to Vmp is larger. Reliability and adaptability of the charge ability then means I can still charge from solar whatever happens appart from the solar panels all breaking together. With MPPT in higher voltage strings if the controller fails you have a lot of head scratching and fixing to resolve.

Charging from a generator I am using a Cisco 2400W power supply (£25) wired into a 1500W boost controller (£18) but the power supply will not switch on if the generator output is too unstable. My generator as I have found out after taking it appart does not seem to have any real AVR so the voltage and frequency are regulated together based on the idle rpm and with it being a 2 cylinder the resulting output oscillates about 2-3Hz which is too much for my inverter charger to lock on and charge. Each power supply and boost controller can provide a 1kW charge rate, so a reasonably cheap charger and I have another two 4200W units to use separately.

I have since found out that I can actually charge the pack from a lower voltage with the boost controller as it seems to be able to regulate the current rather well so my 50V output from the power supply will happily (has not melted or started smoking... yet) step down and charge at 47V. Fuel efficincy is around 20% on a kWh in to kWh out basis.

You have to charge the pack at a slower rate. The cells are all on the market because either the internal resistance has increased too high for fast charging or they have been in an accident (short or fire) and electrolyte has leaked out and the cells then have internal losses and high self discharge. None of the cells on the market will be capable of charging at 400A without a lot of heat being generated, like 20-30%. Plus, the efficincy of the cells at a slow charge rate is over 98%, compared to burning them charging at 400A and wasting 20-30% of the power.

With a boat, shore power, wind, solar, generator in priority order. Charging from shore 24 hours a day at just 500W still adds up to 12kWh per day and only needs a very thin cable. Slow and steady as the tortoise.... not the hare,
Reply
#43
I've opened another thread for my sailboat battery. There is some overlap with this thread but not 100% .. be best to keep separate.. Once I've done the boat .. would like to do 48V home set up so be useful to keep these distinct.

https://secondlifestorage.com/t-LTO-for-...3#pid40823
Reply
#44
(11-14-2018, 09:23 PM)completelycharged Wrote: I now have a 36kWh LTO setup split into 2kWh packs, each with a separate 100A breaker.

Sounds like a nice setup , did your use electrical grease on your connections ? Aluminum has a problem with expanding and
 contracting with oxidizing . The connections will fail ,It is only a matter of time without the grease to keep the O-2 out.
 Aluminum oxide doesn't conduct electricity . I have seen a housing development in Niagara Falls Ontario , 1 out of 3 houses
 burn down because of the aluminum wiring .

Connecting copper to aluminum have different expansion and continuity rates.

https://www.bluesea.com/resources/108/El..._Materials

https://www.alanwire.com/technical-data/...sion-chart
Reply
#45
Well that was a nice diversion while I was away, some very interesting and useful factors to add to the knowledge bank, thanks gents.

So the system has been mostly installed for a month now, still a little fettling and fine tuning to go but for the most part we have been running without issue.

Adter the first week we had to tear down the pack, re-arrange and reassemble to compensate for not testing resistance an AH of the cells and packs. We did measure the idle voltage of the individual cells as we assemble them, the packs were assembled only with cells that were between 2.15 and 2.18. The cells arrived in March and have sat untouched since then and who knows how long before we received them.

Out of the 215 cells about a dozen were at or near 0v, these were recharged to 2.8v and have sat since, after about a week they were still above 2.1v but I haven't checked them since. These will be assigned to packs destined for high abuse, tractor start batteries, 4wd 2nd battery, camp power packs, etc.

The 20s8p packs happily run at AGM settings on the inverter however we have changed to user configurable settings to drop the cutoff voltage to the inverter's minimum of 40v. The bulk of the packs usable power is between 44-48v, which sit about right with all of the theory, 2.2-2.4v per cell.

After checking the specs, menu and settings on the inverter and running some maths a couple of times to be sure, we have decided to add 2 more 8p packs in series to take advantage of the full voltage sweep of the inverter which will charge up to 58.5v, charge voltage of 2.65v per cell.

This means we will have 8p22s 30AH LTO cells, somewhere around the 13kWh mark.

We chose the Sacolar Sunrino MLP 12kw inverter. MLP=MPPT, Low Frequency, Pure Sinewave.

Our first test went a little sideways, connected to mains power the system charged the battery pack, when we switched off the mains power the inverter kicked in and successfully ran from battery, power save mode worked, switching the inverter off, then powering up slightly every 30 seconds to check for load, although it works it is not something we will use. The inverter also worked in normal on mode, however when we switched mains back on a software bug didn't allow the mosfets enough time to change from inverter mode to charge mode and caused a cascade failure exploding all of the mosfets on the +ive rail. From new inverter to paper weight in 24 hours  Confused.

Sacolar were fantastic about the issue and with a little back and forward confirm the situation and settings, they were able to replicate and identify the issue, fixed the software and air freight us a whole new mainboard and inverter/charger board. Provided were detailed installation instructions.

Sacolar have also been very helpful with our request for firmware tweaks and adjustments, the main request we had was to disable the auto frequency sensing for the inverter and be able to lock the output at 50hz, they had a new firmware for us in a week.

Once we had the inverter re-assembled we tested the system with a lab power supply instead of battery pack, so that we were able to simulate and control how the system behaved for testing.

Intriguingly the inverter has a parasitic draw of 3.5A regardless of battery voltage.

Real world offgrid testing has gone well, we have been able to run any device we see fit, as well as day to day household items. So far we have run radial arm saws, bandsaws, angle grinders, welders, water pumps, 2 x fridges, front load washing machine. The combination of Washing machine and water pump equate to 3,700w during the initial fill and heat cycle, drops to 2,500w once full and just heating.

The 150a BMS didn't last long in series as it dramatically limited and charging capacity and occasionally going in to overload protection, which granted is it's job but it was purchased for a much smaller setup.

When we rearranged the pack's for better balancing we moved the BMS to be in parallel, balance only, no over current, over voltage or low voltage protection. Other than a dramatic beep.

This made everything behave better for the chargers and inverters but the balancing wasn't up to the job of such a big system, so as we do, I pulled it apart for some tweaking.

The balancing connection had 2 x 10 ohm SMD resistors in series, I laid some 5A fuse wire across the 2nd resistor to bypass it. This improved the balancing and the heat was not exceptional. The next step will be to replace the 2nd resistor with fuse wire and stack the 2 resistors for parallel connection.

Charging the system from the genset has been without issue, we have a Chinesium 8KVA AVR genset.

We've tested various settings and found the 50A charge to be most reliable ~2500w, when the genset is running the inverter goes into load bypass so that any load is drawn from the genset, with nothing running we could get up to 70 or 80A charging but as soon as any load came on the genset bogged down and/or stalled. At 50A we get reasonably fast charging and no issues with loads. the Frequency wanders mostly between 47-57hz, when the batteries are full and minimal load the genset goes up to 60-62hz which makes the inverter switch output through auto sensing, hence the firmware patch.

Strangely if the output is 60hz the Mitsubishi fridge runs constantly due to the thermostat trigger but does not produce any cold, switching to 50hz instantly produces cold.

The only time we have hit low voltage cutoff was when a water pump pressure switch failed and run for 12 hours, followed by 2 overcast days producing minimal solar input. The were no water leaks so the pump proceeded to eat its own impeller, then continue spinning like a blender full of spare change.

Over the Christmas break we plan to fine tune the solar panels and add a few more, bringing us to a total of ~10kw, through summer this is excessive but through winter we hope to require minimal genset top ups.

I wish everyone a Merry Christmas and a safe New Year
mblowes and Ibiza like this post
Reply
#46
With 22s I do not charge above 56V (2.55V) and do not go below 46V (2.09V). I lose access to around 10% of the full capacity but I then do not have issues with balancing caused by cell capacity differences when you try to go above 2.6V or below 2.05V. This has allowed me to run without any cell balancing for over 6 months now.

I was aware of the inverter switching risk / cascade fireworks issue on older inverter boards as they can switch back to mains out of phase and end up trying to switch 180 degrees if your unlucky with a resulting cascade. I'm surprised that a new inverter had this issue, but not in some regards as it only shows under load on some types of load (motors). What load did you have conected and running when you had the failure ?

You need to connect a scope up to input and output to see the phase alignment to make sure it is switchiong over correctly because a bench power supply alone will show you next to nothing and they may not have fixed anything....

47-62Hz with the genset seems like the governor (sticking) or fuel system (needle valve spring) has an issue.
mblowes likes this post
Reply
#47
Thanks CC,
10v is still a pretty good sweep, once we get the extra packs are built I'll try your settings and see how they play out.

Our inverter was the first of a new model, the serial number is literally xxxx0001. Sacolar have been really good to deal with and stated that they used the same management interface between their HF inverters and LF inverters, they just didn't give the transformers, caps and mosfets enough time to discharge and resync the phase when changing from out to in. The system has been faultless for a month with the new mainboard and firmware.

The cascade failure happened with no load, nothing connected at all, this was possibly just a test case they hadn't tried, with no load there was nothing to discharge the circuit.

The generator has been on the to-do list for a while, it was used and abused when we got it, covered in weird oily soot, brushes worn out, but the price was right: $100 with key and remote start. I have adjusted the governor screw so that it's close to 50hz most of the time and doesn't bog down and stall under normal loads but I have been meaning to check all of the related governor springs, pivots, etc.

Between kids, work and an over ambitious amount of hobbies and projects, there just isn't enough hours in the day.
Reply
#48
(12-23-2018, 12:09 AM)BaronVonChickenPants Wrote: Thanks CC,
10v is still a pretty good sweep, once we get the extra packs are built I'll try your settings and see how they play out.

Our inverter was the first of a new model, the serial number is literally xxxx0001. Sacolar have been really good to deal with and stated that they used the same management interface between their HF inverters and LF inverters, they just didn't give the transformers, caps and mosfets enough time to discharge and resync the phase when changing from out to in. The system has been faultless for a month with the new mainboard and firmware.

The cascade failure happened with no load, nothing connected at all, this was possibly just a test case they hadn't tried, with no load there was nothing to discharge the circuit.

The generator has been on the to-do list for a while, it was used and abused when we got it, covered in weird oily soot, brushes worn out, but the price was right: $100 with key and remote start. I have adjusted the governor screw so that it's close to 50hz most of the time and doesn't bog down and stall under normal loads but I have been meaning to check all of the related governor springs, pivots, etc.

Between kids, work and an over ambitious amount of hobbies and projects, there just isn't enough hours in the day.

I'm trying to design a system to prevent power lose to my house (~15kW max) during short brown-outs.  I will have a LPG standby generator for longer blackouts, but was wondering if the LTO inverter solution might do the trick.  I'm surprised there isn't an off-the-shelf solution for this, as I would assume a lot of people will want this.  For what it's worth, I'm in the import/export business so if you need help sourcing from China, please feel free to hit me up.
Reply
#49
Hi Alex,
The system we have installed would work for what you want to do, however something like the 3 x PIP-5048 in parallel might be easier to work with.

We selected the Low Frequency Transformer based inverter as we do not have mains power, only solar and generator for charging and the PIP units struggle to operate from generators.

The inverter would need to connected either in between your mains feed and house load by an electrician so that it can pickup the load during a brown or black out.

Another option to look at if your not trying to run a whole house or office is some larger UPS units, there are a number of members on the site, myself included, that have added Lithium batteries of some sort to a cheap 2nd hand UPS.


I believe the main reason people are not installing this type of system is simply the expense, they can not justify the cost for the rare inconvenience of being without power. Although how rare this is depends on where you are.

With improvements in solar and battery technology and pricing this is slowly changing to be more viable option and more people are looking to reduce their power bills. 

The number of options available for solar with support for batteries are increasing, but this is not really an "off the shelf" type of product, there is a lot of knowledge required and in most places a licenced electrician is required for at least the high voltage connection of mains and house load.


Thanks for the China sourcing offer, I will keep it in mind for my next order.
Reply


Who read this thread?
100 User(s) read this thread:
hbpowerwall, daromer, Sean, mike, station240, Majorphill, Korishan, A + A, PAF, Franky Beuselinck, Darmok, jdeadman, wim, 1958 greyhound, watts-on, Dan Lim, jesusangel, Gimpy_AK, HyTech, Stefaan De Ridder, Chiptosser, KaminoReal, KtB, Mazlem, Geek, Rad, floydR, hookemdevils22, Gerard Achternaam, DCkiwi, emerca, ChrisD5710, cristof, CarelHassink, karmi27, Solexx X, Ste Collister, Charly144, djuro, Calin, rkelly, BaronVonChickenPants, Rookie, spinningmagnets, Stijn Wever, uploader18, Jason Morris, brwainer, Roland Domke, nrm21, Ahmed_Ali, gregoinc, rtgunner, Drokz, costty, gpn, viralprankster, Vr6chris, Jimbob01, Headrc, rodagaster, Dala, maxime, Crimp Daddy, Redpacket, mldee, EnergyDoc, completelycharged, Ibiza, Onizuka, Zvonko000, iomagico, colibri1, Bomboodle, farmerjohn, indigo, Michel, Greg Verrall, Bubba, JDish, Willowvale, Cherry67, CBlack, Scrud, Sholphin02, pyrophantom, Riplash, 100fuegos, chris330, Jim Jr., bukvaahmed@gmail.com, Katarn_89, w0067814, cprhodin, Kickngas, blackrock, michaell, Gregor Scott, unclebob, Tech4U

Forum Jump:


Users browsing this thread: 1 Guest(s)