Share →

We have drawn no small interest in our SELFISH SOLAR initiative – indicating everybody knows there is something wrong with the devolving net metering situation across the land. And this week’s video takes a deep dive into Selfish Solar. It may get a little technical. So set your Life Alerts on STUN and cinch up your Depends old guys. We’re going in….

As you can see from the recent AmerenUE bill I used 1140 kWh in this period. That normally would have billed $94.96. But I offset it with solar and indeed generated 1597 kWh for a 457kWh surplus. But at $0.0247 per kWh “avoided production cost” that generated an income of $11.29, just barely offsetting my basic connection charge of $11.24. Interestingly, even though I DID offset that connection charge, they still collected taxes on it as if I did not – $1.48 leaving me with a net bill of $1.43.

My point is, this is a rigged game. And it is rigged against YOU. In every way. In large ways. In small ways. Indeed it is the small, sly picayune elements that annoy me the most. They never miss a trick. Taxes on charges you didn’t incurr????

But beyond that, the bigger point is that the polyanna kumbayah concept of investing $100,000 in your roof and using it to generate income from the utility company never was going to happen. And your attachment to that vision is being used against you.

Rather, I would advocate, just pay the damn $11.24 for a BACKUP connection from the grid, and don’t bother with exporting power at all. Generate your own electricity right where you consume it. Store the excess for use at night or during inclement weather. And tap the grid when you need to during extended periods of bad weather or if your system goes down for parts.

But most of the comments we receive on selfish solar are some variation of “just do this and you would STILL be able to export power to the grid.” That rather misses the point. We’re trying NOT to deal with that at all. You just won’t give up the vision. And they will and do use that to control you.

Battery storage of solar electric makes no economic sense at all at the moment. We are spending huge amounts to demonstrate nonsense here. But that’s where I live. What I know is that at some point it will make very good economic sense to do so. If the technology and tribal knowledge is there to do so.

Everything changes. The world is in a state of constant flux. Here’s some good news. At the end of 2016 lighting consumed 25% of our TOTAL electrical use in the United States. And at that point, 88% of ALL lighting in the U.S. was STILL incandescent bulb. In this video I mentioned that we just bought an LED fixture that gets 133 lumens per watt. Indeed, LED lighting uses about 11.5% of the energy of incandescent for exactly the same light.

And the light has gotten better. I can get LED light now down to 2700K temperature (warmish yellow light) instead of the harsh blue white light the early bulbs gave off. And last week I bought 8 of GE’s latest vibrant color LED bulbs that look very much now like an ordinary 60 watt incandescent bulb for $24.95. That’s a little over $3 a bulb.

And so as late as 2016, only 12% of our lighting had been converted to LED. But the prices are falling and the experience of LED is improving dramatically. And very soon now the adoption of LED light to save on electric bills DOES now make sense and is going to take off like a rocket – exponential growth. And in a period of about a year EVERYONE is going to convert.

That means that 25% of our total national electric usage will be reduced by 88.5% to maybe 3% or 5% over the course of the next year or so.

Well that’s fantastic. Less belching smoke from coal fired plants and as everyone knows, if demand for a product goes down, the price has to go down too doesn’t it?

Not in the surreal land of government/huge corporate monopoly structures it doesn’t. It means drastically reduced revenues in the face of the same fixed costs. And so the PUC’s just have a little private meeting with the utilities and in a non-vote vote simply raise your per kWh rate. And batta boom batta bing, problem solved.

Sure, this will be offset somewhat by the adoption of electric drive for transportation. But I’m going to go out on a limb here and predict that your costs per kWh from the utility company are never going to go DOWN in rate. But they will go up. And potentially dramatically. Change causes the requirement to update things in fixed plant investment. And so it will happen.

Against that backdrop, about half of ALL existing residential solar installations in the U.S. have been installed in a little over the last two year period. While focusing on STormy Daniels and the plight of those piteous children from El Salvadore, the media may have failed to mention one of the most significant paradigm shifts in the last 100 years.

And 97% of those installations are grid-tied with no energy storage capability at all. And for good reason. Battery storage adds 50% to the price of an installation and the concept is you NEVER have to use it. The batteries go bad without ever being used other than the twice annual 5-minute outage. And so no one wants to pay that.

Of course, in the event of hurricane, tornado, flood, or zombie apocalypse, that $70,000 solar installation on your roof cannot actually produce enough power to charge your cell phone because on loss of grid power, it becomes TOTALLY INOPERATIVE. By design of the utility company in ANOTHER unholy deal with Underwriter Laboratories on the very pressing need to prevent injury to utility company repair workers of which NOT ONE EVER has had the slightest tingle from a solar installation to date WORLDWIDE.

I hear all the time about the vaunted Tesla Powerwall. I ordered one three years ago. I ordered one again when it was revised and updated and repriced. Hell I’ve ordered a solar ROOF from Tesla. All to no avail. They’ll let me know when they are shipping to my area. Why hasn’t this happened? After paying a brazillian dollars for Solar City, don’t they WANT to sell solar roofs and powerwalls?

They can’t get them approved. I’ve heard from several people who actually paid the price and took delivery. They eventually got refunds because Tesla cannot “get them approved.” In Australia perhaps. Or Puerto Rico. But not here. It just isn’t happening.

The utility oligarchy has already started the war and it has been raging for several years now. But they are using silencers and muffling their guns hoping you won’t notice until it is won. Which is odd because it was over before it began. They have all the artillery, AND your political structure in their pocket. But they are desperately afraid you’ll find out because sufficiently enraged, there have been set backs in some areas of the country by consumers and solar advocates. They are actually very fearful of this whole green thing. Indeed they are rushing to install huge solar installations of their own in a bid to claim the high ground here. Which is actually a good thing.

But they very much want to keep the serfs on the fiefdom and under control. And so teaching them to read and write is strictly verboten.

Edison’s original vision for his electric lighting system involved installation of a steam engine driven generator at each installation site.
This proved impractical because steam engines are noisy, smelly, require fuel, and are subject to frequent breakdowns. And so central “stations” emerged. Westinghouse used Tesla’s Alternating Current to allow large generation sites and the transmission of electricity long distances by wire.

Cool. But that was all developed in the 1880’s. That’s 130 years ago.

If we forget economics at the individual level, and also throw out the brazillions invested in the existing system, how would you design electric power production today from a clean sheet of paper?

And the thing that jumps out is we already HAVE solved most all the problems. You can use magic rocks on the roof to convert sunlight to electricity NOW and we typically just don’t do much with our roofs. They are kind of useless space required to keep the rain off and the sun OUT.

Sun doesn’t shine at night. Store the energy in batteries. SOME homes in the 1880’s actually ran on batteries if you can believe it.

And so you generate power ON SITE at the same geographic point where you USE it. Batta boom batta bing.

Electric grid? Why not? You can move electricity around. There is NO WAY to properly size a solar installation to use because the loads vary constantly and the irradiance of the sun varies constantly. And obviously, the sun goes down in New York long before it does in Los Angeles.

In an ideal world, the grid can be a balancing mechanism and economic arbitrageur of power economics. But it is extremely ill suited to being an energy storage service.

Battery storage has been problematic because of the short life, high maintenance, and large weight and volume of lead acid batteries. Lithium technology solves all of that but it is 3x the cost. But as we adopt electric vehicles, the availability of remnant batteries becomes obvious and the costs of lithium batteries repurposed to solar energy storage plummets.

Selfish Solar is a way to gain CONTROL over your electric energy future and insulate yourself both from the viscitudes of hurricanes and zombie apocalpses AND the frontal attack from your own utility service.

But in getting down to the where and the how, we have a couple of issues. First, all the producers of solar power management equipment are wrapped around an axle catering to the whims of the utility companies.

Second, they have been mired down in a never ending cycle of adding accomodation for yet another battery type/scheme/thingy.

And third, RETROFITTING grid-tied solar installations is just devilishly difficult in some non-obvious ways. The option to go grid-tied takes you down a very different path than having battery backup. And the central issue there is that battery based systems tend to be DC COUPLED in that the solar panels which produce DC power, and the batteries, which store DC power, are curiously ill matched to our universal 240vac two phase power system that ALL of our appliances and devices depend on.

Grid-tied systems by contrast are almost universally AC COUPLED. In that they convert the solar panel DC into 240vac at the earliest opportunity. And increasingly THAT opportunity is moving to the roof itself with small microinverters on each solar panel. Eventually these will simply move INTO the solar panel assembly itself and many future “solar panels” will simply produce an AC output. The wiring and installation issues all get easier when we can bring 240vac down from the roof already packaged for us and connectible to whatever AC system we have in the house.

So our quandry here at EVTV is how to use EXISTING AC COUPLED grid-tied systems with an incompatible DC storage device using the minimum possible equipment.

Grid tied inverters work by detecting the AC sinusoidal waveform provided by the utility grid. They then produce an output that is locked in both fequency and phase to the grid power system. By bumping the voltage just a few volts, they BECOME the current source for loads and the current flows from the grid-tied inverters to the loads. If they produce exactly what the loads require, the voltage of your house system will be just enough higher than the grid to keep from drawing any current from the grid.

In the event that the loads are GREATER than the power output capability of the grid-tied inverters, the voltage falls slightly. But you have basically unlimited power available from the grid to “hold it up” by providing power (current) into the house from the grid.

But if the grid-tied inverters produce MORE power than the electric dryer and air conditioner need, it simply increases the voltage ever so slightly above the grid regulated level, and your system becomes a SOURCE of current and power TO the grid.

This increase and decrease in voltage gets a little unmanageable to explain, involving phase shifts between the grid-tied inverter and the grid itself, but think of it as simply raising and lowering the voltage, ohms law being much easier to explain than RMS averages and phase angle relationships.

And I suppose it wouldn’t hurt to note that the grid-tied inverters detect the presence of the grid power NOT by power, but by frequency. And that a very small window of about 2 Hz off on frequency means LOSS OF GRID to the inverter.

Why is that important. If you lost the grid, and your power output exceeded the level demanded by the loads, the grid-tied inverters increase the voltage to export power to the grid. If the grid isn’t there, there is no place to export to.. And as the voltage rises, eventually the damage and blowup to the attached load devices and the grid-tied inverter itself ensues. And so normally there is some maximum voltage it will go to before protection algorithms simply shut down the inverter. Basically, the grid-tied inverters HAVE to have a place to “store” or export excess power to.

Enter the off-grid inverter. Since all the grid-tied inverters apparently need are a 60Hz sinusoidal waveform, the idea of using an off-grid inverter to provide an output for the grid-tied inverters to synch to arose about 12 minutes after the grid tied inverter was invented.

But if you take a 300 watt inverter working from a 12 volt battery, and use that to trick a grid-tied inverter into operation, experience indicated that it would work, but a few minutes later the 300 watt inverter blew up, and sometimes set the battery on fire. Then the house. Then the neighborhood.

And so using off grid inverters to energize grid-tied inverters without a grid, became a bit of a solar power faux pas. And that lesson has become part of the tribal lore of solar power. Don’t do it. It won’t work. And it will blow up all your equipment. Including your airconditioner.

I have been suspicious of this for some time. And so after developing our battery control system for TEsla batteries and battery modules, I had the instrumentation necessary to observe power flows in real time. And so we shut off the grid, entirely, and began experimenting with it. Remarkably, we didn’t blow much up. Indeed, within certain constraints it worked very well because the transformer based H-bridge inverters we were using allowed the excess power to basically reverse “rectify” right into the battery as DC.

There ARE a couple of provisos here though. The amount of excess power the off-grid inverter can absorb is limited by the power output rating of the off grid inverter anyway. So the 300 watt inverter was NEVER GOING TO WORK in the first place, simply by power rating. You have to use an off grid inverter capable of handling the maximum difference between what the grid-tied inverters produce and the load demand in BOTH directions.

That is, if the loads demand MORE than the grid-tied inverters can produce, the off-grid inverter has to make up the difference within its maximum power output level obviously. But the reverse is true as well. It must be able to ABSORB the power difference between max grid-tied output and loads within its power rating as well.

The transformer is probably important as well. It is a little old-school today, but it does an excellent job of providing isolation. But like all inductive devices, it also resists changes in CURRENT level. And so it is a bit “spongy” call it or resilient to changes in load and grid-tied output. Think of it as putting a bit of slack in the line.

Another problem is that if the battery gets full, and you disconnect the battery, all bets are off. You have lost your storage device to absorb excesses. And so the battery controller must be able to shut OFF the grid-tied inverters before disconnecting. The safest way to do that is to use our CHARGE ENABLE output to energize a relay connecting the grid-tied inverter outputs to the system. In this way, BEFORE reaching full, the battery can disconnect the grid-tied inverters.

We tested this on two inverters of two different manufacturers and at the 10kw and 20kw level several months ago and were amazed at how well all that worked. But it was using Tesla full packs of 300 to 400vdc. The peak voltage of a 240vac rms waveform is 240 x 1.414 or 339.36 volts peak to peak. Well that is right in the wheel house with our Tesla battery voltage. And so I could readily understand how the power was transferred through the transformer, and using exactly the same H-bridge PWM action used to produce the 60 Hz waveform in the first place, through the switches and into the battery in a kind of back handed rectification.

I advocate strongly for the use of full EV battery packs as solar energy storage. It caresses my sense of elegance that we could just take a pack out of a car, plug it in and use it without a lot of reengineering or processing. But quite beyond that, the current levels at 339vdc or 240vac rms are simply a great match for the wiring we already do. 10kW of power is about 30 amps at 339 vdc. And so 20 kW would be 60 amps. These are all comfortable current levels for ordinary house wiring.

The problem is that we have the vast body of viewers who want to experiment with this using weeny packs of two Tesla Battery Modules at 48v. And they want off the shelf MPPT charge controllers and inverters that are by and large 48vdc at this point. Indeed a 300-400vdc inverter has to be special ordered from China and there really just aren’t any available in the U.S.

But the 48vdc “standard” is not very well standardized, and really not a realistic solution unless you are building Tiny House Nation in Montanna. First, the voltages they operate at are really NOT a good match for the voltages of the TEsla Model S pack. And second, to do 10kW of power requires 208 amperes of current. That is signficant I2R loss as heat and you have to use huge cables. And this can be a bit misleading.

We use large cables all the time in electric vehicles. And runs are very short. But there is a false sense of security here. In an EV, we ARE at higher voltages. And while we might do 500 or even 1000 amperes, it is for six seconds. Not all afternoon. A 10kW CONTINUOUS load is a lot of current over a lot of time and so you get away with nothing here. We’re really talking 4/0 cable for even short runs if you want to install this safely. A cable fire NEXT to a lithium battery is little better than a lithium battery fire directly.

But it is clear I’m losing here. And I get it. Tesla Batteries are now $17,000-$20,000. They weigh 1330 lbs. And it is just easier to deal with the 27 inch long slabs at 56 lbs to build what you want. And while you could put them in series, the equipment to do 300-400vdc is again mostly unobtainium.

The problem is, the inexpensive and commonly available 48vdc product out there is just not well suited to the Tesla Battery Module Voltages.

First and most dangerously, almost all the battery types supported want to charge to 56v or higher. If you charge a 2S1P Tesla Battery Module set to 56 volts you will at a minimum destroy it completely. But you risk fire and explosion.

As well, the low voltage cutoff of most 48vdc inverters is 42-44vdc. At 44vdc the Tesla System is at about 56% state of charge. You’re leaving half the capacity of the battery on the table. A 20% level of state of charge would be down around 38.25 volts. And 36.5 volts to actually cutoff operation would be perfectly reasonable.

So we went to a Chinese inverter manufacturer who makes a 48vdc inverter we like and asked them if we bought a small quantity of these would they add a battery type catering to Tesla Battery Module voltages. And they did. We just got them in and began testing and were delighted to find that they worked perfectly with our Tesla Battery Modules.

The unit is 175 lbs and uses a large copper transformer and some good switching electronics. It has three fans that are controlled to come on as needed and the box does an impressive 12kW of continuous power and a whopping 36kW peak output for up to 20 seconds. And the pricing is very good. We can buy them in quantity, ship them here and offer them at a profit at $2995. That’s under 25 cents per watt.

But the really good news is it allows us to start eliminating other devices wholesale. It features an AC input that can be grid or generator. And it will automatically use that input to charge the batteries in the event that battery voltage falls below the battery alarm voltage of about 39v as we measured it. It can charge at up to 120amps. AND it can power the loads while it is doing that. So our 6kW grid tied charger GOES AWAY. The charger is built in.

And so you can use the grid as backup if the battery is drained too low, and indeed you can use a generator. It even has a generator start relay set when it needs to charge. You can use this to start up your generator obviously. And it will wait up to 15 seconds for a stable AC waveform before it starts the charging process. This is rather important when working with generators.

The question then becomes, can we make the MPPT charge controllers go away? And I say controllers plural because as battery sizes and arrays grow, we find the choice and power levels of MPPT charge controllers somewhat limiting. When you get up into the 3kw to 5kw range these things get expensive. And if you want to make use of even a modest 10kw array, you are using two or three of them.

Well there is a way. And it goes to making our system work with EXISTING grid-tied photovoltaic systems anyway. Our Solar Edge grid-tied inverters, for example, use series strings of panels at around 500vdc to make power. 500v is far too high to work with MOST economical MPPT charge controllers which tend to be limited to 100v or 150v at best. We’ve found a couple but they are pricey and a bit erratic frankly when charging to our 50v charge limit.

Many installations today use microinverters on the roof – one per panel that actually produce 240vac down to the equipment room. And so the ability to use AC coupling to existing grid-tie inverters would be huge.

The question is can we do the same charging from grid-tied 240vac with this 48vdc inverter as we did with the high voltage inverters with the full battery pack.

The manufacturer assured us that it absolutely could not and would not work ever under any conditions. It would just shut down automatically and entirely if we tried it.

Well, we ARE located in Missouri. So you kind of have to SHOW ME. I’m not very good with the “being told” thing. In this video, we demonstrate our test results.

One of the things that keeps me going around here is days like that. The results were frankly astonishing. We hit a peak of 280 amperes, 13,500 kw of power IN to our battery through this inverter WHILE the entire shop was running on the Solar Edge output at about 4,500 watts. And we did 230 amperes for 40 minutes until the battery was charged.

The inverter ran cool and quiet without complaint throughout. Never a hiccup.

In surveying this, I would like you to note again the relationship between a 12kW inverter capable of brief sojourns to 36kw, a 25 kw max array output from the grid-tied side, and a battery capable of 1300 amps by design. I don’t think any of this works with a little 5kw inverter.

That said, do note that the cell voltages for the Nissan Leaf and Chevy Volt batteries are essentially the same as the Tesla Model S, 3.6v nominal, 4.2v max charge, and about 3v fully discharged. So voltage wise, the Tesla Battery Module Setting on this inverter would work quite well with those batteries as well. I just ordered a kind of old beat up Nissan Leaf battery pack from eBay for $2300 and we have a new addition to our clan who is working on a controller for the Leaf. A 24 kWh battery probably working at 18kWh at this point is not precisely where we want to go. But it is budget minded at $2300, about the cost of two Model S battery modules.

What remains is to control the Solar Edge output when the battery gets full. We intend to do that with normal contactors on the summed output of the Solar Edge inverters. We’ll use the CHARGE ENABLE output of our battery controller to energize those contactors, and so when we reach the fully charged voltage set in the configuration, the contactors will de-energize and disconnect. When the battery stte of charge falls to the RESUME level, the charge enable output will again close the contactors, and after a 300 second timeout, the inverters will produce again.

Tnere IS another technique I would like to dabble in. Ostensibly both the Solar Edge and Enphase grid-tie inverters have a bit of a hidden feature. If you shift the frequency of the off-grid inverter between 60 and 62 Hz, the inverter produces 100% power at 60 Hz and nothing above 62Hz. But if you could shift it within that window, you can actually modulate the output from 0 to 100% in quite linear fashion.

Unfortunately I have discussed this at length with the inverter manufacturer. Subsequent to our testing and this weeks video they are really QUITE interested in working with us further. But they are literally out of memory space on their current microcontroller. So adding CAN control is kind of out of the question at the moment. But we continue to head scrum on possible future products. They are awakening to where I want to go with this.

Of course, it is a serious disappointment that American made componentry suffers enormous arrogance and hubris sufficient I cannot bear to even discuss it with them. They embarrass me mightily. That is I am embarrassed FOR them. Often relatively small in both size and sales, they have somehow developed the idea that if they ACT like a big stupid corporation, they might get to be one someday. I can only shake my head in wonder and hope they get what they ask for. Punishment enough.

Click on the manual photo below to view the current state of the documentation for this inverter which is now available in our online store.

What’s left here? Well quite a bit actually. We want to perform this testing with Enphase’s new IQ7x microinverter. We’re still waiting on availability for this latest microinverter as it works with large panels of up to 96 cells and was specifically designed for the Panasonic HIT 300 watt panels we already use on our 13,500 watt test array. You might find it curious that Enphase just recently invested $25 million in a 10% stake in SunPower Solar. Recall they make the 22% solar panels and cells that we use on our flexible solar panels. They are going to retire their house microinverter and use the Enphase model exclusively. Long term, I would look to them to do the first viable solar panel with INTEGRATED grid-tie inverter built into it. This would seriously decrease installation costs and complexity.

This week we have begun a new project. We are going to call it the PowerSafe 100. This will be a single cabinet using a very strong heavy Wiegmann NEMA enclosure made of 12 gauge steel. Lifting eye bolts on top and heavy duty wheels on bottom. This monster will be 72x36x30 inches and weigh right around 2000 lbs. But it will hold 20 Tesla Model S battery modules for a slick 100kWh of storage. It will include the 12kw Inverter inside with a liquid cooling system for the batteries, solid copper bars for the high current connections to the inverter, and a control panel mounted in the door sporting our EVIC display and SUB and CAN ports and all manner of switches and dials and controls.

The idea is to get everything in one box and only do AC connections to it. Grid/generator in one port. AC output to your loads panel out one port. And a third port to connect 240vac from the grid tied inverters on the roof. Perhaps a generator start output. This should be enough to power any reasonably sized large home with a couple days storage and the idea is to easily connect to an existing grid-tied solar installation without having to redesign the whole thing from scratch. As we adopt LED lighting and more efficient heating and cooling technology, the storage will actually appear larger and ever more useful.

We see that as a $50k solution probably not attractive to most of our viewers but will demonstrate about the level we think all this will be at in a few years. At 50 cents per watt it is right in line with Tesla’s Power Wall, but you can build it yourself and if you fail to mention to the utility company that you have installed it, I’ll try not to spill the beans from this end. And of course it is 10x the storage capacity and twice the power output.

The enclosure looks a bit to me like a largish gun safe. Ergo the PowerSafe 100 project title. And of course all the piece parts are pretty much already available in our web store if you want to follow along the build yourself. We hope it will BE safe of course but we actually named it after A safe.

I see this as more or less what a future home power station looks like. And I really think it’s possible and workable right now, if perhaps not economically attractive yet to the consumer. If you want to play, we think we can enable it.