Multiple power sources

An isolation transformer from a surplus dealer in Brooklyn

It has been a week since I posted a blog about the need for multiple power inlets on early electric cars.  I have not heard a single product announcement yet.  What is taking so long?  Clearly I am going to have to explain how to do it.

Let's skip over the most obvious case, where you are charging outside in the winter, you have access to two outlets on different circuits at 110 volts 15 amps, and you have a small cabin heater that you bought from J.C. Whitney or the like.  Mount the heater safely, plug it in the second circuit, then your car is warm when you leave and you don't need to run the built in heater as much, extending your range.  If you have air cooled batteries like the Mini E where the air comes from the cabin, then you have even warmed up your batteries and significantly increased your range.  This is as trivial as it is important.

But now we turn to summer.  You have found two outlets that you know are on different circuit breakers, and you want to charge faster than you can from one 110 volt outlet?  First of all, Mini does not want you fooling around with their research car, so forget it.  But if you have a Tesla, a DIY electric car you built yourself, or an electric car you own, maybe you have some options.  The first and most obvious is the Quick 220 safety box.  This device is simple, safe and reliable.  It can detect two circuits that are on different phases and which therefore must be on separate circuits.  It is available in 15 or 20 amp versions and provides a 220 volt outlet, either the type you find on a 220 volt air conditioner, or a twist lock.  Tesla provides adapters for air conditioner receptacles, DIY types will have to make something.  I don't know what the options are for Leaf or Volt owners yet.  Maybe they will be like Mini and stick to the letter of the electrical code, maybe they will be advocates for changing the code to facilitate electric cars in a safe way.  We'll see.

Quick 220 safety box, Model: A220-20D Version 2

The limitation of this approach is that it does not work with Ground Fault Interrupters.  And just about every outdoor or garage outlet in the US has a GFI.  If you find one that does not have a GFI, it should be upgraded.  So what do you do now?  Or let's say you have found two outlets on separate circuits, but they are on the same phase?  The power is there, but cannot be stacked up to 220 volts.  What to do?

The low tech solution for combining power from two GFI outlets into one receptacle involves an isolation transformer in addition to the Quick 220 box above.  (It can also be used for two circuits on the same phase.)  If one of the GFI supplies is isolated and the neutral side of the transformer output is tied to the neutral of other GFI, then the Quick 220 box can safely combine the two 110 volt sources into 220 volts.  There are lots of things you have to get right:  the transformer has to be big enough, the phasing has to be correct, the neutral wiring is tricky.  It can be done, but don't try it at home as they say.  Get a professional to help.  And watch out for cheap Asian or Indian made transformers advertised on the Internet.  Many are not transformers, they are Autoformers.  The difference is that a transformer has electrical isolation between the primary and secondary windings.  An Autoformer does not.  An autoformer will not work here.  Period.

Topaz 91002-11 transformer 2.5 kVA, Dual Standard Electric F246 Powerstat 7.2 kVA combined variable autoformers with T5587 Choke

Once you do get a transformer setup like this working, your car still may not want to charge.  The Mini E is very fussy about the stability of the voltage source for charging, and this is a good thing.  It prevents charging from undersized extension cords, corroded outlets, and other fire hazards.  Nor would it like the voltage drop across a small transformer.  I do not know, but I would assume all electric cars from reputable manufacturers have similar discriminating tastes in electricity supplies.  So what to do?

There are several options if your car charger refuses to continue charging once it realizes the hokey setup you are trying to feed it with.  I have a large Variac that I used to adjust my 108 volt supply at my work place up to 125 volts to charge a little faster.  I expect that could be used, but it is expensive unless you have access to a surplus dealer like I did.  It might be possible to load the transformer with a electric space heater and turn the heater down as the car ramps up its current draw.  But this could cause you to trip a circuit breaker if you don't get the timing right.

All this is way too complicated for the average driver.  The only reason I bring it up is that it can be done.  And the people who should be doing it are the people bringing us the first generation of electric cars.  Modern electronics can do the combining efficiently and safely.  UL might have an aneurysm, and OPEC will probably bribe UL not to approve it.  But it can be done.  And someone should get working on it.

The one thing electric cars need most but no one is offering

Every electric car, no matter the manufacturer (or start up), is offering one and only one power inlet.

Is this the only way it could be?  Gas cars have only one filler pipe for the gas tank.  So electric cars should have only one power receptacle, correct?

Wrong!  At least in the USA.  Much of the world has 220 volts as their standard supply voltage, which makes for a faster charging process.  But in the USA we stayed with 110 volts even after light bulbs switched from carbon filaments to tungsten one hundred years ago.  Since wires are more or less the same diameter in European homes as in US homes, this means there is less power available from US outlets, even if they are nominally safer.

There are many places in the US where it is easy to find multiple electrical outlets near each other that are on separate circuits. I always used two circuits in my garage in winter, one for charging the Mini E and one for keeping the cabin warm, which helps improve the battery performance greatly.  (Of course the charging circuit was usually the 220 volt wall box and the heater was on 110 volts, but bear with me.)

At my work place in the winter, I used two separate, dedicated 110 volt circuits.  One for charging, one for heating the car.  I have done the same thing when stopping to charge at the homes of several other Mini E drivers.  Plug the car into the charging box and run the cabin heater to a regular outdoor receptacle.

I have seen three separate 110 volt outlets at multiple shopping centers that have reserved parking spots for electric cars.  Truck stops with the "Idle Aire" system provide three 20 amp outlets right next to each other, all on separate circuits.  All sorts of camp grounds have multiple circuits side by side.

There is no reason not to provide a second connector, designed into electric cars, so that you can either run the cabin heating or cooling WHILE charging from a separate circuit, or simply double the charging rate if both circuits are 110 volts.

How to protect against someone plugging two cords into the SAME circuit?  This is a very easy engineering task.  If two circuits are out of phase with each other, they are guaranteed to be on different circuits.  This was the case I had at work, this is the case with "Idle Aire", and the shopping center parking lots.

So, why not provide even more than two power inlets on an electric car?  Actually the limit would be three.  In industrial or commercial locations where power is generally supplied in three phases, three would be the maximum that could be definitively distinguished from each other, and thus verified to be on separate circuits.  Idle Aire provides three outlets right next to each other, all on separate phases. Same in the shopping center parking lots.  The two outlets I had access to at work were on different phases of a three phase supply, although I did not get around to finding a third circuit.

In the long run, two (or three) power inlets in an electric car might not be necessary, but in the short run it strikes me as essential for flexibility.  And I don't hear any one talking about providing it.  This is an engineering crime of omission.  I cannot think of another thing that would improve the flexibility and even viability of early electric cars more than multiple power inlets.

And best of all, I don't see anything in the National Electrical Code or recommended practices of the Society of Automotive Engineers that forbids it.

So here is how it would work.

1) Plug in the first cord, the car starts charging.
2) Plug in cord number 2, and if the car can verify that the circuit is separate from the first, you either double your charging rate or use the added power to heat or cool the cabin, as you choose.
3) Plug in cord number 3, and if the car can verify that the circuit is separate from the first, you again increase your charging rate.

I understand that the coming BMW electric car will heat or cool the cabin from "shore power" but only after the battery is fully charged.  With a single 110 volt supply, that is about all you can do.  But with two circuits, it could be extremely useful at times to condition the cabin temperature BEFORE charging is complete.  I speak from experience.  25,000 miles of electric car experience in the Mini E, to be precise.  Much of it in cold weather.

There is a commercially available safety device that can combine two 110 volt outlets into a single 220 volt outlet.  But for practical purposes, it would be of limited use for an electric car although I have heard of a Tesla driver using one effectively.  These devices are popular for, say, contractors that need to operate a 220 volt floor sander in an older house, or operation of European instrumentation in a US facility.  But for outdoor use with an electric car, the omnipresence of Ground Fault Interrupters (GFIs) makes this "Quick 220" box unusable for the average electric car driver.  It only works without a GFI.

However, a properly engineered electronic interface in an electric car could easily take advantage of multiple separate 110 volt circuits to get faster charge times without special infrastructure installations in the early days of electric cars.  And I don't hear anyone talking about it!  Why?  Has no one noticed?  Are you guys asleep or something?  Hello?  Is anybody out there?  Don't make me do it myself!

Why is it difficult to charge from a generator?

A dual section "Variac", 7 kVA maximum.  For use at 110 volt 12 amps, the minimum safe power capacity required is about 2 kVA.

I heard from other Mini E drivers that charging the car from a generator was usually unsuccessful. I never had occasion to try it, but it seems like a valuable backup plan if you could count on using a generator. So I wondered why it doesn't work. The story goes that large diesel generators sometimes work, but the Mini E always refuses to charge from a small gasoline generator.

I believe the reason is actually a safety feature built into the circuitry from AC Propulsion. The car measures the supply voltage before it starts charging, and then if the voltage drops too much as the car starts to draw power, it will stop charging. It figures you are using an extension cord that is too long, too thin, or has corroded connections or is somehow otherwise a fire hazard.  This is why Mini explicitly says not to use an extension cord.

I noticed this safety mechanism once when I was trying to charge from a 110 volt outlet at a considerable distance. I had some very heavy cable (10 gauge, rated at 30 amps) but even though I was only drawing 12 amps, the car would refuse to charge through this very long extension.  So after a long dry spell, I drove on the grass to get closer to the outlet and found that I could charge if I plugged in directly, or even if I used 100 feet of my heavy cord. But at 180 feet, it would refuse. If I recall correctly, the limit was about 8 volts of drop from no load to full load. More than 8 volts drop and the car refused to charge.

I had a large variable transformer for other reasons, so eventually I tried using it to correct for the voltage drop.  It worked, and I was able to charge the car even using 180 feet of heavy cord.  I would set the "Variac" (as it is called) so that the output voltage was equal to the input, which was about 110 volts. Then as the Mini E started drawing current, I would turn up the knob so that the output voltage stayed at 110 volts, more or less. And the car kept charging.

A Variac usually can adjust the output voltage over a range from +20% to -100% of the input

So my theory is that a small generator would have a significant voltage drop under load. A Variac could compensate for this drop. If someone who is still driving the Mini E has a small generator and would like to test this theory, leave a comment and I will get back to you.  Maybe we can try it sometime.

It might not work, since a small generator would also change its frequency with load. I don't know if the Mini E cares about power line frequency, but it might. And frequency is harder to compensate for than voltage is. The large diesel generators that have been reported to sometimes work for charging the Mini E would likely have less variation in both frequency and voltage than a small generator.

If you are curious about the theory behind this voltage drop measurement, see the technical discussion on Wikipedia about Thévenin's theorem.

J1772 standard plug appears for sale on the internet

Those of us who could not renew our lease on the Mini E had the option of keeping the wall mounted EVSE, which is the safety and convenience interface box for charging.  But since it has a connector that is specific to the Mini E, it is of limited use.

Clipper Creek, the manufacturer of the EVSE (electric vehicle service equipment) supplied with the Mini E has not yet been able to supply an upgrade cord and connector to convert to the new standard connector, called the SAE J1772 connector.  This will be used by the Nissan Leaf, Chevy Volt, and other upcoming electric cars for sale in the US market.

Before giving the links to the connectors for sale on the Internet, allow me to speculate as to why UL is giving Clipper Creek a hard time about upgrading existing 240 volt EVSE boxes to J1772.

One of the UL requirements met by the Clipper Creek wall box appears to be that it is explosion proof, like all electrical equipment that might be installed at a gasoline filling station is required to be.  This means that the enclosure is sealed and air tight.  If the cable were changed and the new cable were not exactly the same diameter, it might not meet these requirements.

Even in your garage, this might be important.  Building and electrical codes usually require that all possible sources of ignition in a garage such as electrical outlets (which can cause sparks), switches, water heaters with an open flame, etc., must all be installed at least 18 inches above the floor, where gasoline fumes collect.  Your car does not emit gas fumes, you say?  Not today.  Maybe tomorrow.  Stuff breaks.  It is better if your house does not explode just because a hose clamp splits in your car.

The kind of spark that can be created from a 240 volt 50 amp breaker in an EVSE it much more intense than what you get from a light switch.  Better to keep it sealed in an air tight box, don't you think?  Not to mention that UL requires these EVSE to be safe around cars powered by lead acid batteries that can emit lots of hydrogen gas.  (If you have one of those, code requires that the EVSE starts a ventilation fan before charging in an enclosed space.  The Clipper Creek EVSE can provide for that too.)

So anyone who is considering converting the connector on your Clipper Creek box, please keep the safety points above in mind.  If you think it cannot happen to you, try typing "house explosion" into Google news.

And of course, the current rating on the new cable must equal or exceed the rating of your EVSE.  Don't go putting a 32 amp cable on a 50 amp Clipper Creek box.

Now, here is what I found for sale today:

70 amp connector and socket, 40 foot cable, $475 from Current EV Tech
The picture at the top of this post came from the above link.  The blog where I originally found this catalog said that Current EV Tech once offered a UL certified J1772 plug from ITT Cannon for over $800.  I don't know if this link is also from ITT cannon, which is a very reputable name in connectors.

32 amp plug only, $450
Not much info on this page, it lists a part number that appears to be of Chinese origin. It might be the same as the one above for all I can tell, but no mention is made of a cable that I can find today.

Like all things electronic, I expect the prices to drop and UL certification to be more common.  Just watch out for the seal going into the EVSE box, try to keep it air tight.

Charging safety

I recently posted some photos of the last long trip we took in the Mini E before we had to return the electric car. This 1000 mile round trip involved charging the car at RV campgrounds. This is technically not allowed by the National Electrical Code, which allows an RV to connect to a 50 amp 240 outlet but does not allow electric cars to charge from them. I generally agree with this regulation, as I am an electrical engineer who is quite concerned with safety.  The old connector design is not particularly safe for frequent use, especially outdoors. RVs are "grandfathered" in, but newer safer connection technology should be required for electric car charging.

Shortly after I posted these photos, Mini emailed a reminder all the lessees that this sort of connection is actually forbidden by the terms of the lease. They said in part "...It is crucial to the safe operation of the vehicle that any equipment not supplied by MINI, never be used to charge or operate your MINI E. This includes any electrical connection adapter for the vehicle, OUC, or wallbox, and also prohibits the use of the OUC or the wallbox in any location not specifically intended for its designed or installed use."

I would not recommend that anyone else do what I did. And I should mention that with one exception, either myself or my wife (who is also an engineer) was constantly monitoring the charging connection and process on this trip. Part of the motivation for trying this was to see if cross country travel could be at all practical in a production electric car. The idea was not to repeatedly take long trips with a research vehicle like the Mini E. My conclusion is that cross country travel is just barely practical if the electric car has AT LEAST a 50 amp BUILT IN charging ability. Note that NONE of the several electric cars soon to come to market offers this. The only exception is the Tesla, which offers up to 80 amps as I understand it. But 80 amps can only be supplied from dedicated EV connections, since the largest commonly available general purpose outlet in the US is 50 amps.

I understand that Tesla has figured out a way to allow their customers to safely charge from the connections available at RV campgrounds, the so called NEMA 14-50 outlets which are also commonly used for electric stoves in the US. But Telsa sells their cars, so they can afford to develop a small travel version of the safety and convenience interface, the so called EVSE box (Electric Vehicle Service Equipment) like our larger Clipper Creek box. BMW is only leasing the Mini E as part of a research project, and we should all respect the constraints this places on what is offered.

But an even larger motivation of posting what we did on our long trip was aimed at the Mini E drivers out there that are doing even riskier things. I wanted to show that you CAN use the Clipper Creek safety interface.  NO ONE should be charging their car without taking advantage of the numerous safety features offered by the Electric Vehicle Service Equipment box, even if you are traveling. Yes, it is big but there is that roof rack available from Mini that I showed in all the photos, without saying one word about it until now. Hint hint!

From the way the news is coming in, this will all be moot soon enough when fast chargers are available in more places. But that will take time. Meanwhile, there will always be people pushing the envelope. The real point of my posting was this: If you are going to break the rules, DON'T CHARGE WITHOUT THE EVSE! ALWAYS USE A SAFETY BOX! Really, Mini is right, please don't break any of the rules. But for those out there who are going to break some anyway, be aware of what is a "mortal sin" and what is a "venial sin", so to say.

And especially, do not ever run 30 amps through the small yellow "occasional use cord". I know there are some who have tried this, but there is a large risk of fire. 12 amps max for the little yellow EVSE!

Four light bulbs = 15,000 miles in the Mini E

There has been an interesting analogy mentioned on a few blogs lately, and I have confirmed the calculation.

The claim is that driving 15,000 miles per year in a Mini E or a similar electric car uses about as much electricity as four 100 watt light bulbs burning all year. You can read Tom's explanation here, near the end of the fifth paragraph. So much for those who say large numbers of electric cars cannot be supported by the present electric grid. Nonsense!

To go further, I am curious how many of those 4 light bulbs I can offset by efficiency measures. So I just counted up all the light bulbs in my house, found out how much electricity this house used last year and got average percentages of residential electricity usage consumed by lighting from a Wikipedia article.

Here is how it works out: I have over one hundred light bulbs in my home, but according to the statistics they are probably only on for an average of 2.3% of the time. My math says this is equivalent to two light bulbs being on all year, one of them at 60 watts and the other at 75 watts. That's roughly 1.2 megawatt hours per year for lighting, or 12% of the total 10 mWh this house used last year.

Even so, since about 93 of my lights are now high efficiency fluorescent bulbs, I am saving about the equivalent of one of those 100 watt incandescent bulbs burning all year. One down, three to go to offset the electricity used by an electric car.

We have only been in our new house for several months, but our electricity usage is trending 10% below last year, even though this summer was much hotter. This is because we keep the thermostat at 78 degrees F in summer and open windows at night if is cool outside. So I figure that saves the equivalent of about another two light bulbs burning all year at 100 watts each. Now I am three down, one to go.

Many other electric car drivers have mentioned that they have installed solar power on their homes, which easily offsets more than all the energy used by their electric car. We have looked into solar power, but here in North Carolina we see a better return from installing a ground source heat pump to replace the upstairs zone of our air conditioning and furnace. So we are doing that now.
I have not run the numbers yet, but I feel confident that we have already offset more than enough electricity to power an electric car. We don't have the Mini E anymore, since they would not let us bring it to North Carolina. But we will probably get another electric car in the near future.

Plus, we will probably add solar power in the coming years. We are holding off for several reasons: While the federal and state tax incentives in North Carolina cover two thirds of solar power installation cost, the market for the production credits has dried up here. I understand that in New Jersey, the Solar Renewable Energy Credits (SRECs) can be worth about 60 cents per kilowatt hour generated. But at the moment there is almost no market for SRECs in NC. Plus, electricity in NC costs about half of what it costs in NJ. So payback times for solar power are longer in NC. In addition we are in an urban setting and have a lot of trees, not to mention a hip roof, all of which is less than ideal for solar power. Still, we believe we can design a good system to generate 3 kilowatts of solar power in the coming years. Maybe more over time if micro inverters that support battery backup systems become available. And of course prices of solar panels should continue to drop.

So in summary the argument that electric cars just move pollution from the tail pipe to the smoke stack is ridiculous. It is easy to offset the electricity used by an electric car.

Pictures from our 1000 mile trip

It has been a couple months since we moved out of the support area around New York city, and had to give back the electric Mini E. But before we left, we took a drive to North Carolina from New Jersey in the electric. That post is here. Now that we are settled in after the move, here are some pictures for the big trip in the Mini E:

This was our first charging stop at Ken's house, charging at 50 amps with Ken's Mini E in the background. (Ken has since returned his Mini E as well, but he still has home made electric Metro I bet.)

The next stop is one of our favorites, Bar Harbor Marina and RV park in Abingdon, Maryland. They are friendly, the fee was $10 (shows as a dump station visit on the receipt), the place is pleasant and shady, and best of all their electric service is well maintained. We had charged at 50 amps at this particular parking spot last winter for three hours with no trouble. This time it was summer and perhaps because it was warmer, the breaker did trip after an hour of charging. No problem, we pulled into an adjacent spot where we completed charging without further interruption. And, the maintenance guy came over to work on the outlet and breaker box as soon as he heard we had trouble. His question about the Mini E was unique, he wanted to know if it could be towed behind an RV and charge the Mini E battery by Regenerative braking while it was being towed. (I wonder what Mini would say to that...)

Our next stop was just as friendly, same $10 fee, at much larger RV park near Washington DC. But we had much more trouble with keeping the breakers on at 50 amps. Ken told me he had charged here also, but perhaps it was in cooler weather. In any case, after plugging in and starting to charge at 50 amps, we heard a sound you might describe as "Snap, crackle and pop" coming from the breaker box for 10 seconds or so. Something was heating up and perhaps boiling off moisture. We noticed this at many campgrounds. It was early in the season, perhaps there was a lot of condensation in the boxes from winter. But the car did not charge for more than twenty minutes before the breaker tripped, and the breaker felt quite warm. This is caused either by wires that are not tight, corrosion on contacts, very old breakers that have tripped many times, or under sized wire. In the picture above, you can see that we tried an adjacent outlet without moving the car, an unexpected benefit of the very long six gauge cable that I added to the Clipper Creek box. In the end we moved to an adjacent camp site that was more in the shade, where we found an outlet that did not trip. This was a frustrating stop, but I did not complain. We did not know at this point that we would have this problem at many more campgrounds.

There were many good stops, some where they asked us what we thought we should pay, and we never had a single breaker trip. We even stopped at one place where they did not want to take more than $2 to charge the car. The stop in Amelia's Court House, Virginia, was just plain delightful. At other stops they insisted on full price like an overnight RV stay (usually around $35) even though we expected to leave in 3 hours, but there was no breaker that would stay on for more than 15 minutes. So much for high price meaning high quality.

One very friendly KOA where we stayed overnight was an interesting situation. They had mostly the older hookups, which are only 120 volts. Usually there are two of the common household outlets (NEMA 5-20) and one Travel Trailer connection, a 30 amp 120 volt outlet. We had called ahead and asked if we could use one of their few 50 amp 240 volt sites for a couple hours and then rent a cabin to stay over night. With the charging problems at our previous stops that day, we arrived later than expected and the 50 amp sites were all taken. (I could have put down a deposit but I would have had to pay the full over night rate for an RV.) We rented a cabin and I did what I had hoped not to do, I plugged into the TT-30 outlet and charged the car at 30 amps, 120 volts overnight. That is fast enough to reach 100% by morning, but the Clipper Creek equipment (shown above) is not set up for 120 volts at 30 amp, so I had to use a direct connection. I have since figured out how to adapt the large Clipper Creek box to 120 volt input, by using an international 240 to 120 volt travel transformer to keep the brains happy but wiring the relay for 120 volt operation. Too late now, I don't have the car to test it anymore! But that night I charged without the safety equipment, shame on me. I have heard of someone charging through the small yellow Clipper Creek box (shown below) at 30 amps, which is literally risking a fire especially in warmer weather.

Instead, I wired the heavy orange cable from the large Clipper Creek box directly to a TT-30 plug and connected to the Mini E without the safety box that keeps the car plug de-energized when it is disconnected. Fortunately no children were up and about by the time we plugged in. It turned out fine, but I won't do that again.

By the way, a county park just west of Washington DC had the newest electrical connections we saw. In the photo above, the NEMA 14-50 is on the left and the TT-30 is on the right. We had very good luck charging here. The fee was more than the usual $10 but less an over night RV stay. At this point we were happy to pay, since we were on the return trip and our previous stop had such bad wiring that we gave up with less than 50% charge after paying $35 for a so called 50 amp outlet that would not even deliver 30 amps without tripping the breaker.

While talking to many campground owners, several of them clearly were thinking this might be the Next Big Thing, and talked about adding charging spots specifically for electric cars. Others, as I have noted above, were completely clueless. I hope someone creates an internet site for rating campgrounds in terms of being friendly to electric cars, having good power that can sustain 50 amps without interruptions, etc.

We miss the Mini E, and none of the soon-to-be-available electric car options get close to that built-in 50 amp charging ability which makes cross country travel possible, if leisurely. It is fun to see that Li-ion Motors, near us here in North Carolina, won part of the X prize. They offer a converted Mini on their web site. I have no idea what the charge time is, they have not answered my email yet. But my daily commute is down from 120 + miles to about 20, so I can bicycle a couple times a week. Maybe I will survive without an electric car for now. Sniff.

This is a little off topic, but in Southern New Jersey the power company PSE & G has been putting up solar power panels on telephone poles. There are hundreds of them it seems. Above you can see one of them right in front of Ken's house, where we charged up many times. These panels use a micro inverter. We have looked into adding solar panels on our house, but we have lots of trees around and traditional solar panels do not like any shade on any panel in an array. A small amount of shade can cause a huge drop in power output because of the way the cells are wired in series. Now that micro inverters are competitive, the partial shade problem is significantly reduced. The only remaining problem is that micro inverters are not yet compatible with battery backup systems. And I for one do not like the idea of spending tens of thousands of dollars on a solar power array that stops producing if the grid goes down. Which is what they do without battery backup. It is a safety requirement, technically called Island Protection mode.