Share →

Hurricane Florence again pointed up both the vulnerability of our grid system and the advantages of solar power – along with its’ principle weakness.

Duke Energy is an interesting case.  This utility grid operator is putting in solar photovoltaics at a furious pace all over the southeast.  

The largest is at the Warsaw Solar Facility being constructed in Duplin County North Carolina  — a 65 MW project, but there’s also “the 40 MW Elm City Solar Facility being built in Wilson County and the 23 MW Fayetteville Solar project being constructed in Bladen County.”

It has not been so long since Florida’s Duke Energy announced plans to expand the company’s solar footprint. Its plan is to add up to 500 MW of utility-scale solar in Florida by 2024.

Their efforts have actually made North Carolina the second largest state with regards to installed solar, after California of course.

At the same time, they have been pernicious in fighting manfully to prevent residential customers from installing solar power at all.

This dichotomy is particularly marked in the aftermath of last month’s Hurricane Florence.  One nuclear power plant was shut down entirely and two others operated at limited capacity, while over half of Duke Energy’s 3.4 million customers were without power for nearly a week and some 80,000 suffered outages for nearly two weeks.

But the damage to solar installations across the state was trivial with a handful of panels damaged.

If we take the costs of the national average kiloWatt-hour of electrical power, we find that less than 30% actually goes to power production.  Another 30% goes to transmission, that is the transportation from a power plant to an urban area.  And about 40% goes to distribution, the local wiring within a community.

And of course, both transmission and distribution are still done primarilly with overhead wires that also feature efficiency losses of 10-12%.

This is why I’m increasingly focused on the concept of generating power AT the point of use.  Solar photovoltaic technology is no longer a science project.

Over half of North Carolina’s solar plant has been installed in the last 24 months and indeed you will find this remarkable state of affairs holds true across the country.  But 97% of it has NO power storage capability at all.

To quote East Kentucky U.S. Congressman Thomas Massie, “The world doesn’t need a better solar panel.  It needs a better battery.”

So peering into the future, let me sum up what we are about and why we are about it.

  1. The transmission and distribution of electricity represents 60% of the cost of electrical power and is based on a concept over a century old.  Electricity should be generated and used at the same geographic location. Point-of-use generation.
  2. The extant corporate utility-grid structure does not favor this change.  They will employ any method to prevent attack on the status quo whether or not it makes any sense.  Their favorites are regulatory and legislative obstacles and of course their “nuclear option” is simply to shut off your power.
  3. They no longer go to the expense to do a good job of maintaining their system in the face of weather events or other natural disasters.  It is now “acceptable” to allow customers to be without electrical power for 1-3 weeks. And these events now occur several times per year.  This is the new normal.
  4. Photovoltaic power generation has become economically feasible. But it suffers one fatal – defect.  The planet rotates on its axis and we cannot stop it without causing numerous unintended consequences. No sun at night.

All problems have solutions and usually the simplest solution is the best one.  Lithium batteries last for years, store a relatively large amount of power in a small space, and require very little maintenance.

Lithium batteries are expensive.  But the electric vehicle adoption curve has COMMENCED it’s vertical climb.  Tesla produced 80400 electric cars in the last three months and ALL automotive manufacturers have now announced electric models.

An interesting aspect to this is that consumers are very suspicious of batteries in general, even batteries included. And so to sell electric automobiles, unusually gracious warranties must be provided specifically on the batteries.

This leads to an unusual seam in the zone.  Remaindered batteries from electric vehicle wrecks have zero value in the automotive replacement parts market.

And so we focus on developing the tools to allow you to access the salvaged batteries for use as solar energy storage.

To deal with the utility grid is very simple – simply cut the cord and go off-grid.  But this has some problems of its own.

Everyone knows that the solar energy available varies with the season.  But few realize how MUCH it varies.  We can produce 3.5x MORE electricity here in Missouri in June than we can in December from the same installed hardware.

And smaller weather events also pose a variable.  Four days of rain cuts our power to the bone.  Four days of sun and we have more than we can store or use.

And so for the present, we advocate maintaining a link to the grid.  But we advocate you remain under the utility grid radar by simply maintaining a basic 200A 240vac link to the grid and paying the monthly minimum connection fee.  Do NOT seek a “net metering” agreement or do anything to alert them that you are anything but a very low usage residential or commercial customer.

As it is QUITE common to have empty houses or apartments or dark warehouses for months or years, they really don’t have much in the way of tools to DETECT that you don’t use much electricity – as long as you pay that $11.24 connection fee every month.  And you quite likely need to have a grid connection just for local code compliance.

And so you should have one.  And use it as little as possible as a backup to augment your power system.  

So we’ve inverted the usual scenario.  Instead of a battery back up system for grid power, we envision a grid backed up battery system.

And we advocate abandoning the kumbayah vision of everybody sharing and loving each other. Do NOT seek net metering or compensation from the grid operator.

We call this SELFISH SOLAR.  And of course the end mission remains the same – total independence from the grid ultimately.  And we further see it as based on repurposed EV batteries for solar energy storage.

Selfish Solar

And of course a further mission is to reduce the complexity and cost of all this and primarily the complexity.  How to do it with FEWER components and fewer changes to the vast majority of the extant solar installations in the country – 97% of which are grid-tied inverter or microinverter systems.

And we are whittling it down bit by bit.  We found what is essentially a 12kw bidirectional inverter WITH an auxilliary input for recharging from the grid while remaining isolated from it completely.

And we continue to work on a control system for managing repurposed EV batteries to feed it.

I’ve a new diagram fairly accurately depicting our EVTV shop system.

As you can see, we’ve basically simply connected our PowerSafe100 to a subpanel using a 100A circuit breaker.  This is basically all you need to do to convert an existing grid-tied solar system to run entirely on batteries.

Well not quite all.  You can see we don’t actually have a grid input to the PowerSafe100 at all.  We have a cord we can plug into a 240vac socket on the wall to recharge if necessary. And so if we get low, we manually shut down our inverter, go to the grid meter base and turn on the grid power, and plug in to charge.

That could certainly bear some further automation. But as it is now it is VERY safe and the occasions where we need to do that are mercifully few so far. We’ll see how December goes.

The second thing is control over the grid tied inverters.  Recall we actually do have a switchbox on the Powersafe to connect grid tied solar and control it with a contactor.  But we’ll probably never use it  with our Solar Edge system.

For one thing, its in the other room.  And for a second, I don’t think we’ll need to.  I have an inverter coming with a set of dry contacts allowing me to switch operation from 60Hz to 62.5Hz.  And my thinking is that if the batteries become too full to store excess grid-tied inverter output, we will kick that on for about a half a second.

Which should be just long enough to drive the grid-tied inverters crazy and cause them to drop offline. The rest of the devices in the shop absolutely will not notice.

Grid-tied inverters ALL apparently share the same feature, no doubt mandated by the UL1741 anti-islanding requirement.  They have a 300 second time delay between detecting a valid power waveform and frequency and actually starting to produce power.

And so with a half second bump to 62.5 Hz, I think we can shut off grid-tied inverters of any manufacture, and they will automatically come back up in five minutes.  We would use power from the batteries for that five minutes and so for a few minutes, we CAN absorb excess power from the photovoltaics.  If cell voltaged again reach our cutoff point, we bump the frequency again for a half second, resetting all the grid tied inverters for another five minutes.

This gets us OUT of reverse engineering the Power Line Communications protocol of 300 different grid-tied inverter schemes. We simply derail them any and all with a momentary frequency shift. The EVTV stumble step.  A frequency stumble.

Collin and I have been underwater for several weeks working on the new ESP32 based Tesla Battery Module controller and associated Raspberry Pi display.  They are coming along famously.

Click on the image at right to view an early cut at a User Manual for this fascinating device.

Our most recent work has to do with adding a connection to the Amazon IoT Core Services. This is their public Message Queing Telemetry Transport protocol MQTT broker. They have a useful overview at

The topic of MQTT and our Raspberry Pi Display is the particular subject of our video last week.  Presented herein for your viewing pleasure and convenience.

As described in the video, we basically decoupled the display and the controller and connected each to the Amazon cloud.

This REQUIRES an Internet connection to a Wireless Router wherever the display is located AND the same wherever the Tesla Battery Module controller is located.  I’m pretty comfortable with that as Wireless access to the Internet has become more or less ubiquitous at this point.

For off-grid installations, the cheapest smallest Wifi router can still act as the hub for UDP connection of the display to the controller.  But the benefits of global access to your solar system data will be lost.

Here’s a little cutie I just can’t get over.  Do you know what a wireless router cost in the mid 1990’s?  This one is $24.95 and has open source software in it (OpenWRT).   It’s about 200 hundred times as fast as the $5000 things we saw in 1995.

It weights 1.4 ounces and fits in your pocket.  Requires 5vdc power.  It’s a James Bond style wireless hub.

To do this MQTT magic, we relied on an AWS_IOT library developed by ExploreEmbedded for the ESP32.  Unfortunately, the library suffers from a few serious issues and hasn’t been updated for two years.  So Collin forked it on his github account at and made some timing changes drastically reducing the occurrence of publishing errors and enlarging the size of MQTT packets.  If you have an interest in doing MQTT, this library is a must and it contains links to some excellent ExploreEmbedded tutorials on the topic of setting up Amazon Iot Core and using the library.  For software developers, we of course have our EVTV ESP32 CANdue board available in the online store NOW and obviously it works with all of this because that is what we are using to develop our battery controller.

Meanwhile, Tesla remains the most volatile stock I’ve ever seen.  A dream for day traders and a nightmare for investors.  I’m kind of in Day Trader mode and was buying with both hands this morning as the stock plummeted to $252 per share.

Musk has become a victim of the same press and fake news as Donald Trump at this point.  Reality just doesn’t match.  Their energy storage sales are tripling this year and actually building OR buying enough individual cells for solar energy storage is getting to be a problem.

They just DOUBLED their best ever quarter in the manufacture of automobiles with an 80,400 units quite beyond what even their fanboyz were projecting.

And the National Highway Safety Administration just came out that Tesla Model 3 is the safest sedan ever built. 

(, Tesla said the National Highway Traffic Safety Administration has not only given the Model 3 sedan a perfect 5-star safety rating in every category and subcategory, but it also “has the lowest probability of injury of all cars the safety agency has ever tested.”

“The agency’s data shows that vehicle occupants are less likely to get seriously hurt in … crashes when in a Model 3 than in any other car,” Tesla said.

The car is selling like hotcakes as new owners give test drives to friends, neighbors, and relatives without ANY advertising expense to the company whatsoever.  How is this going to come out????

So they are doubling sales quarter to quarter, on the world’s safest electric car with a 310 mile range and sports car performance, with TWICE the margin of any other automaker. And they have at least three other viable unrelated businesses taking orders and cash – solar energy storage, solar roofs, and semi trailer trucks.

You do need to be a guru to tell the future here.  (Duh NOT). I”m rooting for the shorts to drive Tesla down to $200 per share SO I CAN BUY MORE.  

We’ve become a nation of the silly and the lame. It doesn’t matter what Trump “tweets” and it doesn’t matter what Musk tweets and it really really doesn’t matter what YOU tweet either I have to tell you.

It matters very much to the world what both of them DO.  And I would note that the future of our world very much depends on what you do too.  Some less so on what you tweet.

An unintended consequence of the global Internet.  We did not foresee the possibility that humanoids would confuse online utterance with reality.  Who knew?