Grid tie inverters and HV battery

jonescg

100 MW
Joined
Aug 7, 2009
Messages
4,213
Location
Perth, Western Australia
Hi All,

A question came up on another forum about testing battery packs. Lots of people will have a 100-400 V battery they would like to load test, but a decent load for a high voltage battery can be hard to come across. So the suggestion was to feed power into the grid using a solar grid-tie inverter.

For reference, a grid-tie inverter is one of these units here, which can be had for cheap, or free when you buy a dozen old solar panels from Gumtree.
GTI.JPG

Theser have an MPPT unit which takes variable voltage and current PV at about 200 to 400 V DC and creates a smooth bus for the inverter to turn it into 240 V AC which is synchronised with the grid.

I want to know if it's possible to supply one of these inverters with a battery voltage of 200 V and get it to push power back on to the grid. Solar panels are self-regulating in a way - if you dead short a solar array you get a big spark, but the currents are typically less than 10 amps. That's not recommended by the way - it's very bad for the photovoltaic cells.

But a battery has no such self-impedance - so I wonder if the MPPT unit would current limit, or if you need a separate current limiting circuit ahead of the MPPT unit?
 
These units check maximum Amps available from the solar array. With batteries connected they tend to blow if you are not limiting Amps during such check. Also a Precharge is helpful to keep them alive.
 
suecy said:
These units check maximum Amps available from the solar array. With batteries connected they tend to blow if you are not limiting Amps during such check. Also a Precharge is helpful to keep them alive.

Cheers - I've since heard that it can, and has been done, but it will export power at the maximum rated power depending on the MPPT unit. So if the MPPT works well from 200 to 400 V up to a maximum of 10 amps, a 200 volt battery will push 2000 W onto the grid. But yes, I suspect a precharge circuit might help keep things under control. Or a giant rheostat :D
 
Hello Jones,

I'm back here on E-S since few month. I have less time for the forums so i sometime take a brake.

Yes it is possible to supply the input with battery instead of PV power.

I actually have a Solar project that will use a bank of 8 Enphase M190 micro inverters (that i got for 80cad$) all connected in parallel to send back to the grid the excess energy from my solar panel.

The goal was to have my solar power not only availlable when AC grid is active like the micro inverter require.. But to also enjoy all the power from my solar panel even if there is power outage. So i am recharging 160Ah of TS LiFePO4 battery in 48V with two Morningstar MPPT charge controller and plan to send back to the grid the excess energy with this array of 8 micro inverters.

Their operating voltage is from 22 to 54VDC si this is perfect for my 15s LiFePO4 of 48.0V nominal. What is also nice is that i will work outside the MPPT range ( above 40V) so these will all work together and not fight each others trying to match their equivalent resistance for best MPPT curve which has been a mao concern about parallelig mppt inverters. I have tested and it work very nicely!! I get 95% efficiency conversion between the DC and the AC output.

My goal is to reduce the maximum possible the moment where the Enphase could send power back to the grid and instead choose the moment where i dump the excess power when the grid need it the most using a PLC. Or to sense the power my home need and to only dump the energy from the inverter when my home consume equal or more than the enphase produce ( about 1.55kW) and this way the AC current never flow back to the grid..
 

Attachments

  • Shéma synoptique du projet solaire version 15 décembre 2019 (Custom).jpg
    Shéma synoptique du projet solaire version 15 décembre 2019 (Custom).jpg
    94.8 KB · Views: 3,257
jonescg said:
A question came up on another forum about testing battery packs. Lots of people will have a 100-400 V battery they would like to load test, but a decent load for a high voltage battery can be hard to come across. So the suggestion was to feed power into the grid using a solar grid-tie inverter.
That works - sometimes. Inverters that use hill-climbing, fixed VOC offset or P+O algorithms work with batteries. Inverters that do full sweeps do not. (Picture what the battery will do as the inverter tries to take the battery voltage to zero volts to see what sort of current it will output.) SMA inverters use P+O so can work. HOWEVER modern SMA inverters are transformerless and so put odd voltage offsets on the battery terminals; use extreme caution.

Note that your battery must be within the inverter's MPPT voltage range to get rated performance out of the inverter. For example, the 5kW SMA inverter needs voltages between 220 and 480 volts DC.
Their operating voltage is from 22 to 54VDC si this is perfect for my 15s LiFePO4 of 48.0V nominal.
Be aware that:
1) That will of course pull your battery down to 22 volts, or will pull it down until the BMS trips. Do not allow this to happen; use a controller to terminate discharge before this point.
2) If you parallel transformerless inverters their lack of isolation may cause "fighting" as both inverters try to set their own operating point. It WILL mean that your system is directly coupled to the grid, and you will need to float both sides of the battery.
My goal is to reduce the maximum possible the moment where the Enphase could send power back to the grid and instead choose the moment where i dump the excess power when the grid need it the most using a PLC. Or to sense the power my home need and to only dump the energy from the inverter when my home consume equal or more than the enphase produce ( about 1.55kW) and this way the AC current never flow back to the grid..
There are now small/cheap inverters that will do all this for you. The super cheap Chinese inverters ($300 or so) will monitor your home's overall current and prevent backfeeding. However, these are not UL rated and should be viewed with great suspicion. The better ones (like the Outbacks) will do all that programmably, and will maintain your battery pack, and give you off-grid capabilities. You can now get a 1500W 48V one for about $900.
 
Thanks you Bill.

I made a video to better explain the experiment. Btw these Enphase are isolated from in to output except for the AC ground that have some high capacitance relative to the DC side.

I saw these chineese inverter /charge controller that control the output to contyrol what they dump to the grid, however these are not UL or CSA certified unfortunatly :( .
Thanks for the outback, i was exactly searching for whych brand can do that for me and be certified at reasonable price.

I also wrote to the company Growatt which seem to have a great line of product at cheap prices and these have an office in LA. I will see what theyr answer is regarding certification.

Here is a video that explain my test with the paralleling of micro inverter and using these with battery:

[youtube]GxSafeeUWB4[/youtube]

Doc
 
Doctorbass said:
My goal is to reduce the maximum possible the moment where the Enphase could send power back to the grid and instead choose the moment where i dump the excess power when the grid need it the most using a PLC. Or to sense the power my home need and to only dump the energy from the inverter when my home consume equal or more than the enphase produce ( about 1.55kW) and this way the AC current never flow back to the grid..

I'd generally caution people from experimenting with grid tying, as there is quite a bit of liability in the play. In any case, the way to "correctly" control the output from a dumb grid-tied inverter is by creating a micro-grid, and then coupling it with the grid without backfeeding. Basically there are battery-powered off-grid inverters that will generate the sine wave necessary for grid-tied inverters to operate. Excess production is then dumped into the battery (built-in charger), and when battery fills up these units will shift the frequency of the wave to either reduce (some solar inverters support that) or completely shut down the grid-tied unit. Same battery-powered inverter will provide coupling via grid/gen input port, which it flips to when battery discharges below certain threshold. SMA Sunny Island inverters support all of that, you can find them for cheap on Ebay atm.
 
Thanks cricketo,
I did not know that there are battery-powered off-grid inverters that generate the sine wave necessary for grid-tied inverters. And that SMA does that.

I saw great bargain of some large SMA inverter at very nice price however these are not split phase so you need two that i guess are connected in sync to operate together out of phase in serie.

I will check for these for sure.

My goad is to have all my 1.5kW of solar panel ( 5x 300W 60 cells) used the most efficiently and to be able to use that power and the one from the recharged battery when there is power outage. Used without the need of extrenal transfer switch. The micro inverter was my first choice however i hate to think having solar power unusable when the grid is off!.. I know that the IQ8 from enphase have a solution but it require to use theyr own battery while having lithium battery and BMS is not a problem for me to get at very great price.....

Also about my 5 solar panel these are divised in two group ( one of 3 and one of 2) to avoid problem of shadow so i am not operating at a starting voltage of 200V+... but at 65Vdc for my 2 solar panel and 98Vdc for my 3 others. so grid tie inverter with dual mppt are rare to find with such low starting voltage !
 
Doctorbass said:
I saw great bargain of some large SMA inverter at very nice price however these are not split phase so you need two that i guess are connected in sync to operate together out of phase in serie.

Correct. You need two of those, or you can also get an auto-transformer (autoformer). Most serious brands will not make an inverter that does split phase out of the box due to uneven loading. Auto-transformer is a passive component and will work with any brand of inverter. There is one made specifically for SMA by Midnite Solar, you can also find one on Ebay. Victron also makes an auto-transformer for use with their single phase inverters.


My goad is to have all my 1.5kW of solar panel ( 5x 300W 60 cells) used the most efficiently and to be able to use that power and the one from the recharged battery when there is power outage. Used without the need of extrenal transfer switch. The micro inverter was my first choice however i hate to think having solar power unusable when the grid is off!..

That's where Sunny Island is nice - there is a built-in transfer switch as well, making grid-coupling easy and safe. And it will work with micro-inverters if you set it up for split-phase operation in one of the two ways I described above.
 
My setup with two Sunny Islands and 7 module Leaf pack. A bit of work in progress...
 
cricketo said:
My setup with two Sunny Islands and 7 module Leaf pack. A bit of work in progress...

Yikes. In your house? Under wooden shelves?

I STRONGLY suggest you move that system to a shed outside. That way a dropped wrench will only burn down the shed, not your house.
 
billvon said:
Yikes. In your house? Under wooden shelves?

I STRONGLY suggest you move that system to a shed outside. That way a dropped wrench will only burn down the shed, not your house.

There is a number of reasons why moving the system elsewhere isn't feasible, at least at this time. But also everything is being wired without anything being exposed, so wrench dropping isn't an issue :)
 
cricketo said:
There is a number of reasons why moving the system elsewhere isn't feasible, at least at this time. But also everything is being wired without anything being exposed, so wrench dropping isn't an issue :)
In the two pictures you posted, one shows an exposed +Vbatt terminal next to a nest of wires, and the other shows exposed, free hanging, electrical tape wrapped HVDC splices dangling in front of the rack. Doesn't get much worse than that. And I am sure we're not seeing the whole system.

It's your house. But keep in mind that another member here, with a much safer setup than you had, lost his garage due to a battery fault. He was very fortunate to lose _only_ his garage. If you lose that part of your house, would you be OK with that? Are you certain you would survive that if you were in the house at the time?
 
billvon said:
In the two pictures you posted, one shows an exposed +Vbatt terminal next to a nest of wires, and the other shows exposed, free hanging, electrical tape wrapped HVDC splices dangling in front of the rack. Doesn't get much worse than that. And I am sure we're not seeing the whole system.

It's your house. But keep in mind that another member here, with a much safer setup than you had, lost his garage due to a battery fault. He was very fortunate to lose _only_ his garage. If you lose that part of your house, would you be OK with that? Are you certain you would survive that if you were in the house at the time?

I appreciate the word of caution. If you were to come over for a beer, you'd feel a lot better about what I'm doing :)
 
Back
Top