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.
I am not sure I understand why self consumed power would be at a different rate. Are you not on a net metering arrangement?

In some states the utility sells you power at the retail rate and buys it at the wholesale rate.

In those cases using it is better than selling it so long as you need to use it as opposed to wasting it.

Using it = retail cost
selling = wholesale cost
wasting = equals no value, usually better to sell it at wholesale but some people will do it to spite the utility company or to go off grid and save on grid fees.

It's not like that everywhere but it isn't that uncommon.

Some like wk057 don't have a choice and can't sell it back but I don't know if he is in the multiple rate use it vs sell it situation. I just remember he said it wasn't worth the trouble or cost to apply to sell to his utility.
 
Went through and re-rendered every graph on wk057.solar with a new color setup. Example here:

2016-03-01.png


Key from my page:
Green - PV power fully utilized (combination of loads and battery charging)
Orange - PV Power at nearly full utilization to offset loads (batteries mostly full, loads close to full PV power)
Red - PV Power Underutilized (batteries mostly full, load total less than incoming PV power)



Visualization for when I'm underutilizing my array. Makes for some interesting graphs on some days.
 
In some states the utility sells you power at the retail rate and buys it at the wholesale rate.

In those cases using it is better than selling it so long as you need to use it as opposed to wasting it.
..........

That is how it works at the end of the relevant period in California, but during the year we get full retail credit for what we generate and that rolls over each month until the end of the relevant period. At that point any credit in kWhrs only gets a cash refund at wholesale rates. I think I have known, but this is a reminder that no matter how we complain in California, we do have a pretty favorable net metering policy. I can see how system designs would be influenced by Net Metering policies.
 
Last edited:
That is how it works at the end of the relevant period in California, but during the year we get full retail credit for what we generate and that rolls over each month until the end of the relevant period. At that point any credit in kWhrs only gets a cash refund at wholesale rates. I think I have known, but this is a reminder that no matter how we complain in California, we do have a pretty favorable net metering policy. I can see how system designs would be influenced by Net Metering policies.

That's a pretty fair system actually.
 
Went through and re-rendered every graph on wk057.solar with a new color setup.

Key from my page:
Green - PV power fully utilized (combination of loads and battery charging)
Orange - PV Power at nearly full utilization to offset loads (batteries mostly full, loads close to full PV power)
Red - PV Power Underutilized (batteries mostly full, load total less than incoming PV power)


Visualization for when I'm underutilizing my array. Makes for some interesting graphs on some days.

Love this idea! The result is very helpful for understanding irregularities in some of the graphs. Would also love to see some stats on the graph showing the amount or percentage of time for each category, and some long-term stats on the same data.
 
Love this idea! The result is very helpful for understanding irregularities in some of the graphs. Would also love to see some stats on the graph showing the amount or percentage of time for each category, and some long-term stats on the same data.

Definitely going to be adding more stats soon. :)

I've been optimizing my database queries a bit to make some of this stuff easier. I now have the graphs on wk057.solar showing lines for the roof and ground arrays. :)

Here's a good example day that makes the different orientations obvious:

2016-02-10.png


More stats soon! :D
 
Definitely going to be adding more stats soon. :)

I've been optimizing my database queries a bit to make some of this stuff easier. I now have the graphs on wk057.solar showing lines for the roof and ground arrays. :)

Here's a good example day that makes the different orientations obvious

Love it! That also looks like a good example day of what you can produce under ideal circumstances, no? Unfortunately, I "must spread some Reputation around before giving it to wk057 again."
 
After finally reaching the end of this thread, I must say that I am very impressed with your personal power plant. Although I'm still having a hard time fathoming your energy consumption, it seems that your system is pretty well sized.

I'm currently grid-tied with no batteries with a 44-panel, 11.8 kW array. This pretty well covers our usage on a yearly basis, with 4 people in a two story house and two Mitsubishi i-MiEVs. I wish I'd have known about SolarEdge when I put the system in, as it would make adding batteries much easier than with Enphase micros.

I do have one question that you may have answered previously, but are there any technical reasons for using the Midnite Classic over the Outback Flexmax? I seem to recall it had to do with the arrays' VOC, but I was wondering if there were any other reasons (charge profile, etc.)? I'm designing a system for a dream house right now (not sure if it will ever get built. I have a tendency to create designs for stuff just in case I do something with them in the future) that consists of 1 Outback Radian and 16 Tesla modules. This is basically a "mini-me" of your system. I was thinking about which charge controller to use, and I thought the Flexmax would be ideal as it would interface with the inverter and the MATE display for whatever that would be worth. I shouldn't need to run too high of a voltage on the PV side as the panels would be on the roof of a single-floor house built to passivhaus standards (no traditional HVAC, but solar radiant floors/walls for heat and straight geothermal forced air for cooling). Overall, the loads are minimal, with an electric clothes dryer and 16-amp EVSE as the largest loads. Since lighting and the fridge/freezer run on 48VDC directly, the inverter may spend most of its time in search mode.

Anyway, just curious on Flexmax vs. Classic overall. Excellent job overall on your system!
 
PV1-

I'm about to put in a 24-panel array using enphase. Solaredge was ~10%-15% more expensive. I realize installing batteries is much easier with solaredge, but if you're grid-tied and net-metering, I'm not sure I understand the value of batteries assuming a non-TOU plan, other than grid failure scenarios? Before I pull the trigger, I'd love to hear your thoughts.
 
PV1-

I'm about to put in a 24-panel array using enphase. Solaredge was ~10%-15% more expensive. I realize installing batteries is much easier with solaredge, but if you're grid-tied and net-metering, I'm not sure I understand the value of batteries assuming a non-TOU plan, other than grid failure scenarios? Before I pull the trigger, I'd love to hear your thoughts.

My thoughts on grid-tied solar in general are basically that net metering is unsustainable. Eventually it won't exist anymore, and by doing a grid-tied system basing ROI estimates on numbers that count on net metering you're essentially just gambling that net metering will last long enough for your system to be profitable.

I made a pretty big bet against net metering with my setup.

For your particular situation, you wouldn't be able to use batteries at all with a micro-inverter setup, as far as I know, until someone comes out with a type of grid-tied battery unit that doesn't need solar or inverters or anything at all. Each panel basically becomes its own grid tied system in this case. I don't think I would do a micro-inverter setup unless it were really the only choice (such as an odd layout of panels in various different orientations and/or with widely varying shading across the array).

But to answer your direct question, batteries are pretty useless with grid-tie + net metering without TOU rates, except backup power... in which case, honestly, a generator would be much more cost effective.
 
Went through and re-rendered every graph on wk057.solar with a new color setup. Example here:

2016-03-01.png


Key from my page:
Green - PV power fully utilized (combination of loads and battery charging)
Orange - PV Power at nearly full utilization to offset loads (batteries mostly full, loads close to full PV power)
Red - PV Power Underutilized (batteries mostly full, load total less than incoming PV power)



Visualization for when I'm underutilizing my array. Makes for some interesting graphs on some days.

Your choice of colors obviously reveals your current greatest concern is loss of investment in PV energy due to insufficient storage during the excess electricity periods. Interestingly, many would swap your green and red colors and call green the state of successfully being independent and having enough, and red the warning sign that you're dipping too much into your batteries. But using battery power is meant to happen. So, I can see the argument that something that is meant to happen should never be red. I still think there's a certain mindset that thinks unutilized PV power is a bad thing; these days, we'd rather sell it back to the grid, or add more batteries, because we're PV-starved and even more storage-starved. But in a couple of decades, we'll consider a PV array with 100% utilization "undersized", unless we find truly helpful dump loads (like desalinating sea water and misting redwood forests and refilling aquifers we've depleted over the centuries -- projects that first require we find ways to desalinate in an ecological way, but I believe we can do that) or our storage gets so cheap we can just use it as a bank account, saving up for that next Mars or Moon vacation (escape velocity).

The AI theory that ASI will take over sort of means that there's only going to be very brief periods of contention where all-or-nothing energy use for computational and AI purposes would be useful, and by that time, we won't need to be pondering this, as the ASI's can figure it out. So I actually think finding eternal dump loads is ... well, I was going to say harder than it sounds, until I put it in those words, and it made me think on a geological scale all of a sudden, and now I realize we could dump back into the Earth to keep the core warm, since we're so busy trying to cool it off with geothermal energy production. Just imagine: using your excess solar power to run compressors to cool the Earth's atmosphere while warming the ground underneath, to help "give back to Earth". Hahaha ... but wait --- unutilized solar power turns into environmental heat, right? So, it simply becomes a question of where you dump the heat. Uhoh, I'm starting to take your side of the fence, wk057: you are bleeding the Earth, making its atmosphere too hot. How dare you. You OUGHT to be dumping that heat into the ground, not the atmosphere. I'd need environmental scientists to theorize about whether I'm correct or not. Just brainstorming here.

Now I skimmed through a few months of output on your array, and there's a lot of red. You're right; batteries would definitely still help :)
 
Last edited:
Your choice of colors obviously reveals your current greatest concern is loss of investment in PV energy due to insufficient storage during the excess electricity periods. Interestingly, many would swap your green and red colors and call green the state of successfully being independent and having enough, and red the warning sign that you're dipping too much into your batteries. But using battery power is meant to happen. So, I can see the argument that something that is meant to happen should never be red. I still think there's a certain mindset that thinks unutilized PV power is a bad thing; these days, we'd rather sell it back to the grid, or add more batteries, because we're PV-starved. But in a couple of decades, we'll consider a PV array with 100% utilization "undersized", unless we find truly helpful dump loads (like desalinating sea water and misting redwood forests and refilling aquifers we've depleted over the centuries -- projects that first require we find ways to desalinate in an ecological way, but I believe we can do that).

Good way to look at it I suppose.

I'll point out that I don't reasonably expect this project to ever really recover its investment. It may, and has a decent chance to do so, but the odds are that it just won't work out that way. Guess we'll see. Definitely a bonus if it does! :) That said, I'm not too concerned about ROI on this project. It was more for fun, to stick it to the utility a little, and something I've always wanted to do in general.

Mainly with the under utilization, I just wish I had a way to do *something* with this power that was useful and didn't require a ridiculous investment to do in the first place. Surely every day will start out in the green working to recover the batteries while offsetting loads. It's just sad that so much power will go to waste, mainly during the sprint/summer/fall months.

I may be adding more storage, but that still won't help with under utilization, really. It'll just put some of it to good use on strings of bad weather days.
 
Going off-grid has been one of my main goals for a while now. I'm halfway there in that I generate enough kWhs over a year to cover my usage, though at night I'm still putting electrons moved by a combo of coal and nat. gas in my car's battery.

While outages are less common now than they were just a year ago in my area, we still get nuisance flickers that last long enough to take out my internet connection and reset the clocks. Plus, PA has started to nibble away at net-metering, reducing the allowed solar capacity from 50 kW to 125% of the annual usage. We don't have any TOU rates here, but solar payback for end-of-year excess is nil, maybe 2 cents/kWh while they charge 10 cents for usage.

Eventually, net-metering is going to go away and end up costing solar owners big. This is already happening in Nevada. Having batteries gives one the ability to ditch the grid (or at least being a consumer-generator) and escape the penalties. Net metering really only gives you access to rent virtual-battery storage. You pay a monthly access fee (meter fee). Any energy you put in you get back out at no cost, but pulling more than you generate costs. Being similar to a cloud-type service, rates change, fees increase, and what is preventing the service from disappearing in the future? At least with your own batteries, your only financial obligations are initial costs and ongoing maintenance, and the only thing to stop you from using batteries are when they fail. The utility and government can't realistically deny you access to energy you create and store in your own battery bank. I think that is the motive behind wanting a battery-based system, the sense of energy independence. Of course, as WK found out last month, that sense of independence comes with some real world benefits (like not knowing that the grid was out because it isn't necessary 100% of the time).

Even with micro-inverters, adding battery capability is doable with an AC-coupled system. While not as efficient or graceful as a DC system, a bi-directional inverter and battery bank is added between your breaker panels and the meter so that the inverter can keep the micros running without grid power. The problem with these systems are that solar production is either on or off. Either the micro inverters are outputting, powering the loads and sinking any extra in the batteries, or they are off. The main inverter shifts the frequency of the AC legs up if the battery is near full to shut down the solar production. This puts a lot of un-necessary cycles on the battery pack. But this is all when operating without the grid. When the grid is present and in use, it acts just like a standard grid-tied system.

AC-Coupled is good for solar owners that already have micro-inverters and don't want to change them out. However, for a new system where you think you might want batteries in the future, either a central inverter or a setup like SolarEdge's might be better. My system is at least ground-mounted, so I only need a ladder to reach some of them from underneath, making future changes easy.

Batteryless gives you the quickest ROI. Battery-based gives you the most flexibility. But, if batteries save you multiple fridgefuls of groceries from going bad during extended outages, your ROI increases quickly.
 
Last edited:
Another thing to note about under utilizing PV.

When the panels are in full use, their temperature hovers barely above outdoor air temperature. This is very apparent if you look at a thermal image of my roof. The panels actually keep the roof cool since that energy is being converted to electricity vs being absorbed and heating the roof.

Here is a FLIR shot from during the install of my panels the day after this one section of panels was installed and brought online.

roof-flir.jpg


The temperature outside was about 74F. You'll see the panels are roughly air temp, while my roof is over 60 degrees hotter at 136F.

However, when the PV is not utilized or under utilized, this energy still has to go somewhere. Where does it go? It heats the panels.

A concern of mine is that over the years this relatively rapid heating effect that happens once my batteries are full is going to eventually result in damage or degradation of the panels themselves due to stress. So, I may eventually setup a dump load in the form of a large resistive heater and blower that just heats outdoor air using the excess power, probably using a bunch of small few kW coils so that I can activate however many I need to match production, preferably directly from the DC side of things. This is presuming I don't find some better use of the power.
 
I think geothermal warming would be an ecologically better approach, but I could be wrong, because of all the activity and material needed to set up facilities to move electricity into underground heat.

There isn't some industrial process that's energy intensive that is material light and relatively nontoxic and doesn't care if you suddenly stop and restart you could set up in a shed and make some $ on? Of course, that sounds a lot like a battery, only more complex, throwing business and economy in the storage container.

Wait: resistive heating? You could cram an element in the bottom of a long metal tube, insulate the top of the element so it doesn't conduct heat back up, and hammer-drill that baby 100 feet down, hopefully with some non-conductive mid-section conduit (not sure how to achieve that; metal might be easiest), and of course wires to deliver the electricity. No big geothermal wells and heat pumps needed! Why didn't I think of that?

The metal conduit can double as a ground rod. How much heat loss would be conducted 100 feet up out of the ground vs. directed into the ground around the pipe? Perhaps it has to be 200 feet deep ...

I must be reading too much space exploration junk. They have a Mars robot that will hammer 100 feet down. In reality, 100 feet is a long way ... at least 30 feet would be something.
 
Last edited:
Well, I could probably setup the resistive heater dump load for under $1k, and that's if I use the DC side to do it (thicker wire due to lower voltage). If I use the AC side I could probably do it for a couple hundred bucks, but then I'm utilizing my inverters just to make waste heat, which isn't really my favorite plan.

There's nothing else I can think of for dumping up to ~45kW that can be done for less $ per watt.
 
Does the property-line utility legislation cover rotary converters? You'd have a monster electric motor in your basement that runs an underground driveshaft, spinning a generator in your neighbor's basement. Or, wireless power transmission (if they have the technology for that much of a gap)? I hate seeing energy production go to waste (once had the grid go out during prime solar production. That was painful), but I can't think of anything that would use 40 kW on-demand unless you had another pair of Model S parked and ready with low charge.

I'm kind of thinking a motorized tarp on a spring-loaded roll that covers the ground array when nearing a full charge. Make it white or mirrored, so that the energy that would've been absorbed by the panels is instead reflected back out to space, keeping the planet that micro-fraction of a degree cooler. It'd keep your panels cool, too.

I still find that very intriguing that solar panels actually stay cooler under load than being idle. They also work better when they are cooler. Consider my mind blown.
 
Mainly with the under utilization, I just wish I had a way to do *something* with this power that was useful and didn't require a ridiculous investment to do in the first place. Surely every day will start out in the green working to recover the batteries while offsetting loads. It's just sad that so much power will go to waste, mainly during the sprint/summer/fall months.
Convert it to hydrogen and store/sell it ?