Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Plan: Off grid solar with a Model S battery pack at the heart

This site may earn commission on affiliate links.
A lot of people buy for "cool" factor or because they're misled about incremental savings vs air source, but in our climate I don't see GSHP justifying under 4000 sf. When the tax credits go away I can't imagine what will happen to that industry.

With the GSHP's the COP ratings are also not the only part of the story. Pumping all that water around takes a significant amount of energy. If you're in an area that allows well-to-well pumping, you might have a 2 or 3hp pump down there running full time. That eats some kWh.
 
Just passed my first MWh of AC side power utilized from the solar setup/batteries. :)

Total system efficiency is looking a bit better than I had originally figured (around 90% from the panels to the batteries to the loads if I'm doing the math right), but I'll wait a bit longer so I can ensure I have enough data for my numbers to be accurate.
 
Nice. That's not far off what wimpy grid tie setups on 'green' homes run in a lifetime. I don't understand why more people don't shoot for a brute force method. I suppose it's kinda mandatory if you want to off grid charge a pair of electric cars, but it sure works.

Do you have any datalogging on your battery SOC over time? I'd be interested to see something like a SOC graph over a week or so period. Did you do any low SOC testing, loading your little grid down hard at say 20% SOC? What kind of sag do you get, and how do the inverters handle it? If you have incandescent lighting somewhere, can you notice dimming while switching on an inductive load, some kind of large motor?
 
how is work comming along on your "dump loads"? once you nearly double the number of panels (with the groundmount panels), and the fact that peak production season is comming soon, you'll probably run into the battery full situation, and need to dump the excess energy...
 
Nice. That's not far off what wimpy grid tie setups on 'green' homes run in a lifetime. I don't understand why more people don't shoot for a brute force method. I suppose it's kinda mandatory if you want to off grid charge a pair of electric cars, but it sure works.

Do you have any datalogging on your battery SOC over time? I'd be interested to see something like a SOC graph over a week or so period. Did you do any low SOC testing, loading your little grid down hard at say 20% SOC? What kind of sag do you get, and how do the inverters handle it? If you have incandescent lighting somewhere, can you notice dimming while switching on an inductive load, some kind of large motor?

I'm logging everything at 5 second intervals. :p

Currently don't really have a lot to go on, but I'll post some graphs and such later on.

I did two tests so far where I ran the pack down to about 15% SoC. The AC output voltage on my inverters starts to suffer a bit once the pack gets below about 42V. Still fully usable, though. At 15% SoC the AC output voltage was down to about 215VAC. Low, but fine for normal loads. DC voltage sag is pretty nonexistent. Literally just over 1V of sag at 55kW AC side load. I over sized my wiring and bus bars to minimize DC side losses.

The inverters work pretty well with all loads I have. The one exception is my one aging heat pump/AC unit. When it kicks on the surge briefly loads the inverters to the point where a bunch more will come out of power save mode to compensate, then settle once it's running. They handle it, although a hair less smoothly than the grid would. Honestly, to me this just points out a problem (the AC unit) that I probably wouldn't have noticed otherwise.

how is work comming along on your "dump loads"? once you nearly double the number of panels (with the groundmount panels), and the fact that peak production season is comming soon, you'll probably run into the battery full situation, and need to dump the excess energy...

I actually started some scripting for the HVAC (Nests) to adjust the inside temperature in response to the pack approaching full. Also, dump loads aren't *required* for my setup, but they'll be nice. If the panels are making more power than I'm using and the pack is full it just doesn't use all available power from the panels (wastes it). Somewhat unfortunate, but, such is life trying to be off-grid year round. No way I'll be able to use *all* excess power I make in the summer. Not without having somewhere to store it in another form, long term, even at low efficiency.
 
I'm logging everything at 5 second intervals. :p

Currently don't really have a lot to go on, but I'll post some graphs and such later on.

I did two tests so far where I ran the pack down to about 15% SoC. The AC output voltage on my inverters starts to suffer a bit once the pack gets below about 42V. Still fully usable, though. At 15% SoC the AC output voltage was down to about 215VAC. Low, but fine for normal loads. DC voltage sag is pretty nonexistent. Literally just over 1V of sag at 55kW AC side load. I over sized my wiring and bus bars to minimize DC side losses.

The inverters work pretty well with all loads I have. The one exception is my one aging heat pump/AC unit. When it kicks on the surge briefly loads the inverters to the point where a bunch more will come out of power save mode to compensate, then settle once it's running. They handle it, although a hair less smoothly than the grid would. Honestly, to me this just points out a problem (the AC unit) that I probably wouldn't have noticed otherwise.



I actually started some scripting for the HVAC (Nests) to adjust the inside temperature in response to the pack approaching full. Also, dump loads aren't *required* for my setup, but they'll be nice. If the panels are making more power than I'm using and the pack is full it just doesn't use all available power from the panels (wastes it). Somewhat unfortunate, but, such is life trying to be off-grid year round. No way I'll be able to use *all* excess power I make in the summer. Not without having somewhere to store it in another form, long term, even at low efficiency.

You could dump the excess to a hot water tank or an ac unit. It's just a matter of hooking up the controls. They make a desiccant dehumidifier too that has been used for dump loads on solar hot water systems.
 
No way I'll be able to use *all* excess power I make in the summer. Not without having somewhere to store it in another form, long term, even at low efficiency.

I have no idea on the technical feasibility of this, given that you're off-grid, but is there a way you can switch to selling your excess capacity to the grid when your pack is full and you have nothing else to do with the excess power? Might as well get something for it rather than letting it go to waste.
 
If you really want to dump juice, get a rack of old SHA256 miners.

That's interesting the AC voltage drops that significantly with a low input. I wouldn't expect that kind of behavior and for it to still allow for high loading without fault. So, the input voltage range of the inverters is more problematic than the internal resistance of the pack at low SOC's? I suppose you're running something near 1/4C at a 55kW discharge, so I shouldn't be so worried. Is that rack two full 85 packs, or is it the full proposed 191? Whats the deal with the 191 anyways? Did you ever measure the drop across your wiring and connections on the DC side? Since you're vaguely closer to utility scale here, its interesting to see what losses you get with practical oversizing of the conductors. I'm just wondering how much the juice is worth the squeeze.

Did you ever mess around with varying your output voltage? Obviously theres a lot of different interactions there, but I notice some loads are perfectly happy to draw a little less on a little less voltage. Although some things draw more and it could end up being a wash, and some things will have reduced performance so its kinda equipment dependent. Presumably you're shooting for 240VAC? It's possible the inverters handle hard loads better at a lower or higher voltage as well. Typical A/C units run the scum class of motors, PSC. Technically, with a PSC motor, you know its working if you get a huge current spike on startup. A hard start kit is designed to increase this current spike so the motor starts up faster, minimizing the appearance of lights dimming, and giving the compressor more oomph to overcome higher head pressure from short cycling or unbalanced conditions. I'm not so sure this works off grid when your inverters are effectively generating the power after its needed. They don't know this 200A load is online until the voltage tanks, and they're left reacting. Higher quality inverters deal with this fairly well, but a big motor starting up direct on line, is about as tough as it gets. You can reduce it, with a soft starter, smaller compressors staged, or drink the juice and run BLDC. My inverter mini split is only a 3/4 ton, but it fires up with so much damn grace. No noise, thud, sag, shake, dimming... I almost feel sorry for the grid when my other conventional A/C unit fires up, but all the lights are LED so I don't notice the little blip. My TED sometimes does, though.

Do you plan on keeping a grid connection long term? The obvious choice would be to dump excess on grid. I see grid tie inverters that are too big for residential use go cheap all the time. But you're going to need to dump a lot of juice to overcome that $18 meter charge or whatever it is.
 
You could dump the excess to a hot water tank or an ac unit. It's just a matter of hooking up the controls. They make a desiccant dehumidifier too that has been used for dump loads on solar hot water systems.

I have no idea on the technical feasibility of this, given that you're off-grid, but is there a way you can switch to selling your excess capacity to the grid when your pack is full and you have nothing else to do with the excess power? Might as well get something for it rather than letting it go to waste.

Few ideas come to mind. As for selling to the grid, it'd be a configuration change to make it work from a technical perspective. However, I can't get net metering here since my system is > 20kW. I'd need a PPA, which no one seems to want to get back to me about.

If you really want to dump juice, get a rack of old SHA256 miners.

I have a ton of those... but then the problem becomes what to do with the heat from them.

That's interesting the AC voltage drops that significantly with a low input. I wouldn't expect that kind of behavior and for it to still allow for high loading without fault. So, the input voltage range of the inverters is more problematic than the internal resistance of the pack at low SOC's? I suppose you're running something near 1/4C at a 55kW discharge, so I shouldn't be so worried. Is that rack two full 85 packs, or is it the full proposed 191? Whats the deal with the 191 anyways? Did you ever measure the drop across your wiring and connections on the DC side? Since you're vaguely closer to utility scale here, its interesting to see what losses you get with practical oversizing of the conductors. I'm just wondering how much the juice is worth the squeeze.

Keep in mind that my pack voltage range is below the voltage range the inverters are designed for. I'm at ~43.2V nominal lithium, the inverters are made for 48V nominal lead acid. So, the AC drop below the normal input voltage range is probably to be expected.

My stationary pack is 36 modules from 85 packs, so 191 kWh. Each 85 pack has 16 modules, so, 85/16 = 5.3125 kWh... * 36 = 191.25 kWh. Under absolute max continuous load for my setup, 64kW, and at ~90% efficient I'd be at about 0.37C load. Still nothing really to the cells. The ~1V drop I see is at the inverters due to wiring lengths. The voltage at the cells is virtually unchanged under load.

Did you ever mess around with varying your output voltage? Obviously theres a lot of different interactions there, but I notice some loads are perfectly happy to draw a little less on a little less voltage. Although some things draw more and it could end up being a wash, and some things will have reduced performance so its kinda equipment dependent. Presumably you're shooting for 240VAC? It's possible the inverters handle hard loads better at a lower or higher voltage as well. Typical A/C units run the scum class of motors, PSC. Technically, with a PSC motor, you know its working if you get a huge current spike on startup. A hard start kit is designed to increase this current spike so the motor starts up faster, minimizing the appearance of lights dimming, and giving the compressor more oomph to overcome higher head pressure from short cycling or unbalanced conditions. I'm not so sure this works off grid when your inverters are effectively generating the power after its needed. They don't know this 200A load is online until the voltage tanks, and they're left reacting. Higher quality inverters deal with this fairly well, but a big motor starting up direct on line, is about as tough as it gets. You can reduce it, with a soft starter, smaller compressors staged, or drink the juice and run BLDC. My inverter mini split is only a 3/4 ton, but it fires up with so much damn grace. No noise, thud, sag, shake, dimming... I almost feel sorry for the grid when my other conventional A/C unit fires up, but all the lights are LED so I don't notice the little blip. My TED sometimes does, though.

The AC output voltage is configurable, but it can't hold the setting once the DC voltage drops < ~42VDC. I have it set at 246VAC just to keep things nice and high under normal conditions. As for the AC unit, it only has any really voltage drop effect on the AC side at startup if all of the slave inverters are in power save mode. If I take a few more out of power save before turning on that AC unit, there is virtually no noticeable issue on the AC side when it surges. Problem is keeping more inverters out of power save increases idle usage by ~35W per inverter. It adds up.

Do you plan on keeping a grid connection long term? The obvious choice would be to dump excess on grid. I see grid tie inverters that are too big for residential use go cheap all the time. But you're going to need to dump a lot of juice to overcome that $18 meter charge or whatever it is.

I plan on keeping the grid connected for backup purposes. The inverters can seamlessly phase in power from the grid to run loads in the event that the pack is depleted. Currently I just have it set to hard transfer to the grid if the pack hits ~15% SoC.

I may try to dump excess to the grid, but not if they aren't going to pay me for it. :)
 
Those inverters are ~90% efficient at 64kW? Or does that include all losses? That's so much damn heat for a basement to handle without you noticing. Presumably its not sustained for more than a few hours, but thats a lot more than a little portable AC unit can tolerate. You'd probably be able to get away with an 18K BTU inverter mini split, or closer to a 2 ton conventional. I'd guess a ~1 ton would hold up since presumably those high peaks won't be sustained more than a few hours, but it wouldn't be overkill. Cooling isn't all that hard, mini splits go in really easy if you have a clear path for electric and the lineset. In a basement, it should be a few hours to get one in.

35W is a lot to consume with no benefit. That's nearly 1kWh a day, every day. I worry about 1W. AC can be really elegant at times, but off grid it just sucks to have this equipment running 24/7 so the light on your fridge works while you sleep. If theres a good way to remote wake inverters, it's not hard to have it automated when a thermostat calls for comfort. Presumably you'd want a few second delay, so the condenser itself would be fired up a few seconds after the inverters wake. If theres some sort of enable style pins on those things it would be really easy, as long as they don't disable if its toggled while they're on or something silly.
 
The manufacturer claim is 92% efficient. I came up with 90% by measuring amperage/voltage at the battery pack and comparing input to output. Considering I run at a lower nominal voltage, this seems fine.

Fortunately super high power draws are an edge case for me. Literally the only time I'll see anything close to maximum power draw is if I charge both Model S at 20kW at the same time in the winter when all heat pumps are on with at least one resistive aux heat enabled..... basically a near nonexistent use case, but I like having the capacity and being able to better wear level my equipment.

I do see loads in the ~15kW area regularly, however. HVAC + cooking + hot water + laundry + normal loads, etc.

For cooling I'm going to probably just exhaust that room to the outside when it hits a set temperature. I also have the new heat pump hot water heater (which is awesome, btw) that I'm waiting on the ducting kit for. It will be run to source and vent air from/to the inverter room, which won't hurt.

For idle loads, the inverters have a search function (~5W idle) that waits for loads and immediately wakes upon a voltage drop on the AC side. Unfortunately this is useless for me since I will always have some idle AC load in my house.

My goal is basically to be off grid with minimal compromise. Efficiency improvements and smarter use of available power is fine, but when I want to crank the AC or the heat, watch TV, work on my PC, run my heat gun, whatever... I want that power there under normal circumstances. I'm good with not charging the cars or adjusting the HVAC by a few degrees on stretches of rainy days and such, but for my other various loads probably not much changes happening.

I have some network equipment and a server I run 24/7 (~350W). I generally leave my far from energy efficient desktop PC on 24/7 also (varies, but been averaging about 10-12kWh/day) for immediate availability convenience. Have a 1.6kW pool pump that runs ~10 hours/day in the spring/summer. Several TVs, office equipment, additional PCs, laptops, couple of mini fridges, etc etc etc. Definitely not an energy frugal lifestyle and I don't really have any intentions of making any major lifestyle adjustments in that department. I've swapped out nearly every light that will see any real use to LED equivalents, swapped out the water heater with the heat pump unit, the one heat pump I had to replace is more efficient than it's predecessor, etc. Where I can cut usage without affecting usability/comfort/convenience by much or at all I'm jumping on it, though. :)
 
I wasn't aware those inverters were that inefficient. Not really such a big deal from an overall system perspective, but I could see how the heat can be problematic in some conditions. Personally I would prefer if there was equipment in place to ensure the temps stay in check without venting air. Throwing hot air out the window is great and all, but air must be coming in somewhere else to make up for it. Less critical to indoor air quality when you're talking about venting semi conditioned basement air, but still not as ideal as a balanced system or simply conditioning the existing air.

I'm a big fan on putting the right equipment in place so I don't have to care about how I use it. HVAC shouldn't be such a big load that you even need to be concerned with the thermostat set point on your system scale. Improving your homes thermal efficiency probably makes a lot of sense here. It's going to depend on how bad things are now, but it sounds like theres a ton to gain. If you're willing to go to extreme measures, you should be able to literally cut your heating/cooling needs in half or more. Ductwork leaks can be quite considerable as well, but thats generally tricky to fix right. This home efficiency stuff is great, since it just kinda works without you doing anything, or knowing anything about it, generally for the life of the house. It's typically more labor than raw material cost too. Going around the house, with a few cases of spray foam makes a huge difference. Cooling my house to 66º and 45% RH today, when it was sunny, humid, and ~85, consumed ~4kWh. Thats 3-4x better than from when I moved in, and its still an R13 wall, although I think theres 1" of rigid insulation under the parts that aren't brick. Heating is a whole lot worse, since some days see 80º differentials inside to out, but cooling is trivial since the differential is really quite small. It's the air leakage that kills you on those 80º days. Solar gains can be problematic too, but that has more to do with the house design and orientation.
 
Last edited:
I think the problem with a conditioning approach is that at say, full load, I'll be producing ~7kW of heat in that room. So, I'd need to have equipment that could remove 7kW of heat.... which is going to take a good amount of power. Comparing that to a louvered intake vent on the bottom of the door to the outside combined with a relatively low power high CFM exhaust fan at ceiling level...... I think the exhaust fan will probably win out.

That said, I'm probably going to do both. I'll likely get an efficient mini-split unit that will work for my average and slightly above average use cases, and the exhaust fan/intake vent as a use case for when the split unit can't keep up. (Independent of each other, never both running at the same time.) I'm also planning to better insulate that room thermally and acoustically once I have something in place for climate control like that, too. Right now I can hear the fans from the inverters under load in the living room if there is nothing else on upstairs.

Overall it's going to be an ongoing effort I think.

Edit: Also, in the winter I'm not overly concerned about the waste heat and will probably want some way to spread this around the house. :p
 
Last edited:
7kW is just under 2 tons, or 24,000 BTU. That's big for a mini split, but I don't think you really need a unit big enough to handle that 24/7 since the load should only persist a few hours I'd imagine. You just need enough thermal mass to allow the unit to run flat out, minimizing temp rise, until the load goes away. A 1 ton mini split should be fine, since MOST of the time it will likely be overkill.

The Fujitsu 12rls3h would probably be the one. They make the same series in a 9, 12, and 15K BTU. The 12 is rated 0.79kW full load, cooling. You can get 2 and even 3 ton mini splits, but the efficiency drops way down, costs go up. Fill the room with bricks or water or something to soak up the spikes, I'd imagine a 1 ton unit would be fine.
 
7kW is just under 2 tons, or 24,000 BTU. That's big for a mini split, but I don't think you really need a unit big enough to handle that 24/7 since the load should only persist a few hours I'd imagine. You just need enough thermal mass to allow the unit to run flat out, minimizing temp rise, until the load goes away. A 1 ton mini split should be fine, since MOST of the time it will likely be overkill.

The Fujitsu 12rls3h would probably be the one. They make the same series in a 9, 12, and 15K BTU. The 12 is rated 0.79kW full load, cooling. You can get 2 and even 3 ton mini splits, but the efficiency drops way down, costs go up. Fill the room with bricks or water or something to soak up the spikes, I'd imagine a 1 ton unit would be fine.

Interesting. I wasn't aware they were *that* efficient. Not too bad at all. A water loop with a radiator and a tank isn't a terrible idea for dealing with peak times either. :p

Edit: for that particular unit looks like it wouldn't work with a generic thermostat....... *shrugs*
 
> But you're going to need to dump a lot of juice to overcome that $18 meter charge or whatever it is. [zomgvtek]

WY Net Metering law leaves the monthly 'equipment charge' intact (untouchable) come end-of-year accounting. Does PA operate differently?

Also WY law prohibits home owners, ranchers, businesses from selling electric power to 3rd parties.
--
 
Does it need to work with a normal thermostat? That's not really how those mini splits work. They come with a remote, and they all have better programmable wired thermostat options, but they don't really work like traditional units. I think you can get a box for Mitsubishi's that works with a standard 24V thermostat, but I don't think LG or Fujitsu offer such an item. Mini splits typically hold a few degree window from the setpoint for efficiency. They're more for set it and forget it operation.

Now, if you can tolerate 3/4 ton, the 9RLS3H is rated at 0.50kW, rated cooling load. And they can run higher than rated load short term, but you'd need to manually kick it on or clone the remote and get something to monitor temp and blast some IR. The 9 does 12K peak, the 12 does 13,6K.

- - - Updated - - -

>
WY Net Metering law leaves the monthly 'equipment charge' intact (untouchable) come end-of-year accounting. Does PA operate differently?

Also WY law prohibits home owners, ranchers, businesses from selling electric power to 3rd parties.
--

Do they just give you a net metering meter? I'd imagine the POCO would want you to have a grid power option to give you a net meter, and presumably you'd pay for the one if you use it or not. Otherwise, couldn't you just put one panel on the roof to save the meter costs every month?