NEC Batteries Solar Part 1

Welcome to the Battery, Solar and Interconnection Aspects of the National Electric Code Class! This is a continuation of a previous National Electric Code review.  In that course, we cover code as it relates to onsite energy systems.  In this class, we’ll discuss the portions of electric code specific to batteries, solar, and interconnection, to better understand design principles, with a focus on articles 480, 690, 705, and 706. 

This class cannot legally provide verbatim sections of code or full tables for the viewer. I highly recommend you purchase a copy of the National Electric Code, as it is a resource you will turn to again and again.  

Let’s dive right in!  Start with Section 480 with the definition of battery cells, which are the base unit of the battery . Electrolyte is the fluid between the cathode and anode. Terminals protrude from the cathode and anode, out of the battery, which attach to the battery inverter cables.

Helpful hint: Lead acid is at 2 volts while lithium ion is at 3.6 volts to 3.8 volts. There are different types of both lead acid batteries, and lithium battery chemistries. For example, lithium cobalt has a slightly higher cell voltage (3.6 to 3.8 volts), than lithium iron phosphate. Therefore, when the cells are wired to create a 48 volt battery, the lithium iron phosphate will have a slightly lower operating voltage for the same number of cells than lithium cobalt. 

Section 480 contains the general battery requirements which is then expanded on later in section 706 for higher voltage battery systems. In general, the battery NEC rules are fairly basic and expand as grid-tied batteries become more popular.  Let’s start with the general requirements and then get to the more advanced sections towards the end of the program.

Storage batteries and battery management equipment must be listed, but lead acid batteries do not. Basically, this implies that as the market moves towards lithium-ion batteries, we will see an increased focus on listing requirements. There’s more discussion on that to come. 

Here is a call-out for overcurrent protection not being required on prime movers for 60 volts or less. Prime movers are using the battery to start ignition or digital controls circuits that start up other systems, like an electric start generator. 

Ambiguous writing brings uncertainty into 48V battery bank design. While much of code only applied to higher voltage batteries, why would prime mover circuits be specifically called out here, if the intent of code was to only regulate batteries over the 60V DC threshold? This gives inspectors wide berth to interpret battery systems, subjecting any grid-tied battery system to the code requirements of higher voltage battery banks. 

A potential scenario: An inspector feels like there should be something in a battery circuit like overcurrent protection, even though it isn’t specifically called for within the code itself.   The fact that the battery is under 60V DC might not convince the inspector. We would like to see clarification in future volumes. At any rate, an electrician might find the current national electric code lacking for batteries. Remember:  code establishes a minimum threshold for safety. It is always okay to add additional fuses or disconnects in line with your battery system if you believe they are warranted. 

With regard to the specific lack of overcurrent protection or disconnect requirements for batteries under 60V, I recommend using listed batteries with built-in disconnects and overcurrent protection systems. Then, if  you’re skipping additional disconnects or overcurrent protection, I would refer an inspector to earlier code on feeder and power distribution taps, which do not require such items if made within 20’ of an electrical service. That is to say, if your batteries are within 5’ of the inverter and under 60V,  there’s no code-driven reason to include external disconnects and overcurrent protection.

Having overcurrent protection and disconnects on electrical power supply circuits is a good idea. However, many consumer-grade lithium batteries already have these features built in so adding external over current protection can be redundant.

Basically, there needs to be disconnects to allow higher voltage batteries to be divided into lower voltage batteries for servicing (where serviceable). Still, with the right planning, adding overcurrent protection is not a bad idea, even when skipping an additional disconnect, so that if a fuse does happen to blow, it doesn’t require getting inside the battery pack to service. 

With batteries, where external disconnect requirements come into play are if the batteries are far away from the battery inverters. If there is a wall that divides them, being in adjacent rooms or where the batteries are outside the building and the inverters are inside, where the cojoining cables go through a wall, it needs a disconnect. Item B says that if the disconnecting means is activated via remote control, and where those controls are not within sight of the battery system, the local disconnecting means needs to be locked in the open position to avoid someone remotely shutting it on the service panel.

So generally, common sense disconnecting says you need a disconnect if the two related items are not within sight of each other, and code generally treats that as within 25 feet. The disconnecting means shall be legibly marked with nominal voltage, current rating, and other items you’d expect to be called out on a listed component. Here’s a battery made by Fortress Power with a built-in 250 amp DC breaker built right into the battery pack, which should meet the code requirements of strict AHJs without the need of an additional external switch or overcurrent protection. 

Battery chemistries with a corrosive electrolyte the structure that supports the battery shall be resistant to deteriorating action. Batteries are heavy. When building a battery room, you put them on a shelf on a rack. That shelf or battery rack is usually metal. Obviously, that’s very conductive. I remember one of my first battery projects. We were picking them up with our on-site lift equipment by attaching some metal chains to the metal casings of the batteries. The battery terminals are just sticking out right there and so we had to tape the chains with some rubber resistive tape to move them! Otherwise, we’d get arcing between the battery terminals and the metal chain that could continue to conduct into the metal lift equipment and potentially electrocute the driver!

So, metal racks are not good for batteries without some insulative properties!  Concrete is considered to be conductive if it gets wet, so in damp basements, etc keep this in mind.  Likewise, you could have some voltage creep out of a

battery, and if it is on a metal rack potentially you can have the rack be energized and that be a safety hazard, so use rubber mats if you’re putting the batteries directly onto a concrete slab.

Here is a battery that is designed for floor mounting and it has an insulated hard plastic base so that it can sit right onto that concrete slab. Similarly, you would put a rubber mat on your metal battery rack or maybe an anodized metal is more resistant.  Painting the metal rack alone is not considered insulated battery rack.

Let’s talk about ventilation for a minute. There are sealed batteries and unsealed batteries. Unsealed batteries, like traditional flooded lead acid, produce gas as they’re being charged or discharged and that gas needs to be vented to the outside.  The question becomes, “Does a sealed battery also need to be ventilated?” 

You could say a great advantage of having a sealed battery is that you don’t need that ventilation. But, what happens when a sealed battery becomes unsealed? Let’s say the building catches on fire. The fire makes it to the battery room and starts burning up the casing of the batteries and it heats the battery to the point where it too catches on fire.

In that case, the sealed bat

tery becomes an unsealed battery. and so has been one um industrial battery fire and

(you have time here because I condensed, do you want to explain this fire?)

uh i believe it was arizona’s aps where uh it was actually the battery itself as far i’m not sure the investigation is complete yet but not not many suspects in this fire to begin wit the battery goes through some internal failure it catches on fire and all of a sudden it’s leaking gas into the room and for whatever reason and i don’t know the real details of this so just hypothetically speaking,

If your ventilation is not on an emergency or backup power circuit ,and so your ventilation isn’t working, when the fire is going on you can get gas accumulation inside that room and that’s exactly what happened in this APS fire.

The batteries caught fire, the room was not ventilating, the gas accumulated inside the battery container, and when the fire department showed up as soon as they opened the door the room exploded, sending several firemen to the hospital.  So, what we’re going to see is that as the sealed batteries get larger in size the ventilation requirements become mandatory even with a sealed battery. The takeaway from this is that you can’t just assume the battery room does not need to be ventilated simply because you have a sealed battery.

We will discuss these ventilation requirements in more detail in other parts of fire code, not national electric code. So, we see national electric code referencing the fire code for ventilation of  larger battery banks.

I don’t think you’re wrong on where the right location is for the battery. The best location for a battery is inside, in air conditioning, in a dust free environment.  Just like other solar electronics, you’ll get much better life and performance  that way.  Dust is as bad as temperature or anything else as far as I am concerned.

The main thing is live parts operating at 50 to 1000 volts nominal and this is squarely within the realm of home battery banks and  even  yeah i guess lead acid has 48 volt nominal

The lithium ion industry will refer to any battery bank near 48 volts as a 48 volt battery bank but on the spec sheet you’ll see a 48V battery resting voltage is typically above 51 volts. 

When you walk into a battery room and see exposed terminals, unprotected cables, and metal busbars outside of traditional distribution panel enclosures, the customer might ask if they touch those terminals, am i going to shock myself, and the answer is yes.

Either the batteries are bought individually and then wired together, or the next step up is a manufactured system where all the exposed parts are protected, or it’s something in between the two where the installer uses the battery manufacturer’s product and the inverter manufacturer’s product and you still have to wire them together.

That means you might go and buy the cheapest battery on the market and put it on a shelf in your garage, but if has exposed terminals you’re going to run into more design constraints than if those terminals are enclosed in a case provided by the manufacturer.

When needing to protect the user from exposed parts, that does not necessarily mean a locked room that can only be accessed by qualified persons.  It means you have to have a locked area, or a barrier, or a locked cage inside a garage to prevent people from getting into area where the batteries are. So per code, battery systems can be in a garage, even when on a shelf using different equipment from different manufacturers, if the proper barriers are considered.

Just as common is to buy a higher end manufacturer assembled system where the provided hardware protects already protected.?? Other suggestions are to elevate the batteries above the floor. Eight feet is the typical height to steer clear of mixed use areas. In the harshest circumstance, an inspector is entitled to ask for 30” from the equipment be unreserved space, such as where a vehicle would not be allowed to park.

These are standard work area rules found at the beginning of NEC. Less than that, a tough inspector may require this dedicate space regardless of how listed or protected the equipment is. Then you have to treat it like a locked cage.  So basically, if you want to put your batteries in the garage and you’re buying uh kind of just the battery and not something more top shelf you’re going to have to cage it off.

Do it yourselfers be warned: you can’t just buy a homemade battery with exposed terminals and keep it on a shelf.  You’re liable for that if you’re going to build your own battery pack and keep it under 50 volts and even just playing around with it with DIY, you need to secure the immediate location and make sure that secured area is a dedicated space. 

 You need to have a 1” clearance between the cell container and any wall. We’re going to get into differences for a completely enclosed battery inverter system like the Tesla Powerwall.  The Powerwall has its own battery and inverter completely enclosed in one shell that can be mounted directly to the wall. 

A battery that’s more like a car battery, which is just a big block with exposed terminals coming out of it, and you are allowed to build your own battery assembly. It’s a process that’s been going on for years and years, pre-Tesla era, and there needs to be a 1” clearance between the cell container and the wall. 

So, you’ll see battery mounting brackets that will separate the wall from the battery when you are wall mounting your batteries or if you’re floor mounting the batteries, you can’t just butt them up against the wall.  They’re not permitted contact with that adjacent wall. Even the shelf itself will have this clearance requirement and we’ll reference that a little bit later. 

There. the terminal clearances how much vertical space you need between the battery and the shelf above it. That is something that’s driven by the manufacturer spec so that’s not a clearly defined vertical clearance code. If you are putting your batteries in a battery room, you want that door to open outward rather than inward.  

Once your batteries get over specific voltage, and we’ll see this in the battery section more advanced battery section the doors need to have panic hardware so that you can’t get locked in to a battery room, even though it needs to be a walkable door.  Gas piping is not permitted in dedicated battery rooms! Keep that in mind with your electric water heaters.  

Another one that is a little it should seem obvious, but illumination is required in your battery rooms. I did a project where we put the batteries outside the building but the home was on a hillside, and we had a little alcove where we could put the batteries, tucking them up against the building foundation.

 Well, that little dedicated space if it’s not just a box that you’re standing outside of and then working into, if it’s an actual battery room, that battery room needs lighting. So, if you’re building a battery room adjacent to the building, it can’t just be four walls and a ceiling.  It needs to be a little bit more professional.  

All of your working clearance requirements apply that you see in the earlier sections of code, such as vertical spacing and clearance behind you. Basically, in a dedicated battery room, you need to be able to fully stand up in that room. That’s 6.5’ of vertical clearance and anywhere from 3-4’ behind you before you get to the next wall. Battery rooms have to be rather sizable and professional. If you don’t have enough space for a battery room, build a battery box instead.  Don’t build something that’s halfway a room, and don’t build a battery box that is so large that it’s really a room. 

In this adjacent section to 480 with batteries we have capacitors. In most national electric code in your battery circuits, when you’re sizing your cable you’re looking at the ampacity of the battery circuit and then you’re adding 25 additional headroom for that. There’s differences between continuous output that’s running full-time and intermittent output that is maybe running for under three hours or less.

In the beginning parts of code, we see either 100% of the intermittent output, the discontinuous output or an additional 25 for continuous load with capacitors it’s a special thing where they say the conductors have to be 35% rather than 25% and so as a building engineer with solar and batteries becoming more prominent, you may find yourself needing to take a master electrician exam.

For instance, you take all the sizing considerations.  Typically, you’re oversizing your conductors by 25% when it comes to electrical design. There are special exceptions, and solar is one of them.  We’ll talk about why a little bit later.  Capacitors are another one of them. Capacitors get a 35% oversizing and I believe that’s just because capacitors can be so rapidly discharged.  

So, that’s pretty much it for Section 480. It’s a very short section and requirements are being migrated towards the more specialty sections at the back of the book. So, the general rules sections end at Chapter Four and then we get to Chapter Five, we start to get into specialty requirements. Chapters One through Four are to do with services, feeders, wires and cables themselves, then general appliances like air conditioners, lights and dishwashers, etc 

Chapter Five  and onward starts to get into specialty applications, special ocurrences which isn’t within the scope of this class.

With lighting fixtures in a battery room, it’s not a terrible idea to have the lights be battery powered on their own, such as LED light strips glued to the ceiling. That can be a cheap and pragmatic way to add lights to where they don’t currently exist. For a higher end build, having the lights on a dedicated backup circuit is not a bad idea. 

Before we get to Article 690, let’s look at what we’re skipping over. Chapter Five give deals with special locations like boat dock. Chapter Six gets into special equipment like electric vehicle charging. For electric vehicles there are ventilation requirements for the garage based on the amperage ratings of the electric vehicle chargers.

Finally, we are at Article Section 690 for solar systems. Definitions start, and the most confusing definition is functional grounding.

A functionally grounded system is one that’s connected to ground. It’s confusing because it is an umbrella term which refers to grounded and ungrounded DC-side overcurrent protection systems, which have traditionally been connected to ground on one end. This practice results in a blind spot within the ground fault detection system, and the solution was to not attach either the positive or negative end of the DC circuit directly to ground, but instead to fuse both sides of the circuit and leave the circuit “floating”, meaning no direct reference to ground. 

But a floating system still has an equipment ground conductor and has its metal parts connected to ground, independent of the actual DC positive and negative circuit. So both the floating style DC circuit design, and the traditional style where one of the two circuit ends were positively attached to ground, are considered functionally grounded systems. 

Inspectors were concerned about “ungrounding” the DC circuit, and this term was added to assure inspectors that the grounding of metal parts wasn’t being thrown out by changing up the ground fault detection circuit. This blind spot was only a problem for commercial-sized arrays up at the combiner box – we don’t have time in class to cover this issue any further, but Google the Bakersfield or Dietz-Watson fire to find out more information.  Maybe add more here?

Well as a final note on those fires, it wasn’t just a blind spot in the inverter ground fault detection circuit that was at issue – the actual fault that burnt the buildings down was due to a failure to consider thermal expansion of the conduit up on the rooftop. Long exposed conduit runs on the rooftop can pull apart if expansion joints are properly considered, even if the conduit was otherwise properly installed at the beginning of the project.

 Thermal expansion would cause the conduit to pull out of the connectors, and then the sharp conduit edges would bite into the wires contained inside them. When the wires hit the grounded conduit, normally, a ground fault error would occur. Due to the inverter blind spot, the ground fault went undetected, and commercial inverters stopped connecting the DC conductor directly to ground, isolating the ground fault detection, as a result. 

Pro Tip: It’s really not confusing if you accept that all metal parts of a solar array need to be grounded, regardless of whether the inverter is a “grounded inverter” or a “floating” inverter. They’re both functionally grounded. 

The generating capacity of the system, according to National Electric Code, is the inverter output in kilowatts. This is a common question on interconnection agreements. If you’re asked on a permit docent what is the system size is, and it does not specifically ask for the DC size of the solar array, the correct answer is to provide the AC output size of the inverter.

System voltage is another interesting issue. Most believe that residential homes must use a 600V solar array, with commercial being 1000V and utility being 1500V or more. But the 600V residential restriction only applies to systems on or inside the building. Let’s say an off-grid cabin is in the middle of the woods, with an adjacent field 1000’ away from the cabin.

In other words, the solar array is ground mounted and far away from the house. What do utilities do to transmit electricity over long distances is to step up to extremely high voltage because as voltage increases, amperage decreases. Voltage is more sensitive to maxim limits – a 600 volts component works fine at 600V but can pop at just a little bit more, like the bursting of a just a little too full water balloon. Voltage has more to do with the toughness of the cable jacket rather than the thickness of the cable itself. 

Whereas amperage is a slower burn, having more to do with the thickness of the cable itself. If amperage limits are exceeded, the device might not fail immediately. It might run hot and start to smolder before catching on fire. Bringing the topic back home, by doubling the voltage, the amperage is cut in half, and so the required cable thickness is cut in half, and so less material is required to transmit electricity over a very long distance. This is why the power grid uses high voltage for transmission and distribution, with distribution circuits exceeding 10k volts and transmission circuits exceeding 100k volts. 

So if the solar array is really, really far away from the point of use, it isn’t a terrible thought to consider moving to a 1000V platform and three phase electricity, because this equipment is commonly available for commercial applications and code only constrains its use when literally on or inside a 1-2 family dwelling. Commercial buildings prefer 1000V products because of those savings and they don’t have the 600 volt residential constraint. Standard solar cable and most solar panels, by the way, are rated for 1500 volts.

Battery Combining and More

In this webinar, we will discuss a few mounting options, but understand that you also have the flexibility to use the T-channel built into the eFlex battery itself for custom application.  The eFlex was designed around this M6 hammer nut and screw. The hammer nut fits into the two T-slot channels built into the back of each eFlex, to assist mounting to a wall or shelf. If you forget the size of the T-bolt, it can be found in the eFlex installation manual.

The size of the ring terminal for the eFlex can be found in the installation manual and on the specification sheet. The ring terminal size we recommend is 3/8ths.

As far as the cable sizing, each eFlex battery is recommended to be charged at 55A and discharged at 60A. Less frequently, the eFlex can be discharged up to 100A.

eFlex are wired together in parallel, and in parallel, amperage adds up – so for two eFlex in parallel, the recommended discharge rate doubles to 110A with infrequent discharges up to 200A being possible (as well as greater inrush current capability).

During normal system operation, the inverter DC amperage is split between the batteries. For example if two eFlex are on an 8 kW inverter, to determine the DC amperage we divide by the battery DC nominal voltage rating of 51.2V, and so 8000W divided  by 51.2V is 156A.  Divided by two, means that each eFlex will see a maximum continuous amperage is 78A per eFlex. Normally this would be an undersized battery, being out of continuous operating range for the eFlex which could trigger a reduced warranty. So in this configuration, it is important to select a closed loop communication inverter (such as Schneider or SolArk) to keep within the product warranty. Installers may always email for a quick design review.

NEC requires the cable to carry at least 25% more amperage than this (if no other factors call for additional increases in ampacity), and so the cables to each eFlex need to be sized for at least 98A. If combining using a 75C power distribution block, the minimum cable size is #3.

However, it is reasonable to increase the battery cable sizes further, sizing off the battery capability rather than the load. This ensures that if the load ever increases, or if a design mistake occurs by putting to large an inverter on too small a battery, that the conductors are rated appropriately.  The eFlex can deliver 100A per battery, and in the event that one battery is disconnected and the other is working, the battery could deliver 100A. Sizing the cable for 100A using a 75C terminal requires #1 battery cable.

Let’s not get into balance of system material just yet. First, let’s learn a little more about site layout with the eFlex.

First, let’s start with a floor mount. It sounds easy, and it is, but for more than one reason.

The eFlex has a IP65 frame, which means it can take direct water spray, and is sealed against dust and humidity. This does not mean it can be submerged, and the battery terminals need protection from rain as well. But the eFlex can be placed in a dirty, humid, outdoor environment such as a barn or garage and its protective enclosure means you do not need to be as concerned about the durability proofing of the box.

The eFlex can be mounted directly on the floor. Most batteries should not be set on concrete. However, the eFlex has an insulated base that provides protection against conductive wet concrete as well as providing insulation against thermal exchange with the slab. It is a battery for rugged environment.

In extremely cold environments such as Canada it may be necessary to build heated box for the battery. A box may also be built for physical protection in a multi-purpose area. But remember the eFlex is already in a robust case, so the box built around the eFlex can be simple.

Many installers put their eFlex batteries up against the wall. This can be space efficient, but keep in mind that in garage locations, the batteries should not be installed where they can be struck by a car or truck.

The floor mount bracket, included with each eFlex, secures the battery to the wall. This wall mount bracket must be used, because it prevents the battery from accidentally tipping over.

Sometimes it is more desirable to have the eFlex stick out of the wall, which makes the overall layout more compact and reduces battery cable length. But if the eFlex is to stick out from the wall, such as when contained in a box, or even when used outside a box, then how does one use secure the mounting bracket to a wall?

At this point, be creative. A 2×4 could be mounted to the wall, and then solar L-foot plus lag screw left over from the solar array could be used to make the required 90 degree attachment point to land on the floor mount bracket. It only needs to be secure enough to prevent the eFlex from tipping if subject to accidental contact.

When mounting in this orientation, it might make sense to buy angled compression ring terminals. Here is a picture of a ring terminal with a 90 degree bend built into the terminal. This allows the cable to hug the wall in a clean manner, while reducing the cable length.

The only ground mount orientation we do not allow is with the terminals closest to the ground as shown. Be careful – this disallowed orientation also applies when the eFlex is on the wall.

The eFlex Wall mount is another popular option for keeping the eFlex out of the way and protected, as well as reducing the cable sizes as much as possible. The end result is a very professional look. The wall mount kit is an additional accessory, but it is inexpensive and will save money on the battery cable budget. Because the inverter is typically wall mounted, it makes sense to wall mount eFlex batteries.

The eFlex can be mounted on the wall in an upright orientation, but also upside down. It can only mount sideways in one orientation – like the floor mount the orientation with the terminals on the bottom side cannot be used.

When mounting on the wall, the vertical spacing between the wall mount brackets is flexible, only constrained by the distance of the T-slot on the eFlex itself. If the eFlex is mounted on its side, the spacing between wall mount brackets should be the spacing between the eFlex t-channels.

When mounting on a wall, it is easy to transition from the battery into a box. This additional protection is not required by code, but most eFlex installations require multiple eFlex which need to be combined ahead of the inverter. The transition into the box requires a strain relief connector.

Battery cables are commonly located within the same room as the inverter, often less than 5 feet away from the inverter. They are not required to be in conduit – although an inspector may want to see additional physical protection based on the use of the room. Keeping exposed cables short and tidy on a wall will make your inspector happy.

Strain relief takes tension out of the cable which might otherwise stress the battery terminals, resulting in a poor terminal connection or even worse, the cable pulling out of the crimp or terminal creating a safety hazard. When cables transition into a box, a strain relief connector is used to suppress tension in the cable.

Another strain relief strategy is to use fine stranded cable. Fine stranded cable has an additional advantage of having a smaller bending radius than regular stranded cable which allows it to bend inside a box with ease.

The use of 3/8ths compression ring terminal, a strain relief connector, and a durable fine stranded cable is a great way to transition from the eFlex wall mount into a gutter before transitioning to the inverter.

We will talk more about battery combining strategy at the end, but for now let’s look at eFlex layout with two eFlex on the wall and a 48V residential battery inverter. This configuration can easily fit on the back wall of a garage, and only requires a few feet of battery cable. It also looks good and is well protected. In areas prone to flooding, wall mounting the system above flood level will dramatically improve the value of the system to the user.

The last eFlex mounting option is shelf mounting or rack mounting. It’s called both shelf and rack mounting, because the eFlex was designed to fit onto a server rack – but rather than attach directly to the rack, it rests on a shelf that is attached to the rack. The server rack bracket is then used to secure the battery to the shelf.

Server racks are not cheap, but can open up some interesting applications. An open server rack may nearly organize a large number of eFlex in a compact, cost-effective manner. An enclosed server rack costs more, but is a good application to protect the0 eFlex against accidental contact in a multi-purpose room.

A server rack may be small, containing a single eFlex and a small rack mounted inverter to protect a refrigerator, internet equipment, and some LED lights while tucked out of the way on a wall in a kitchen. Larger server racks are used in commercial installations inside shipping containers supplying power to large buildings.

It is not necessary to use a server rack to shelf mount an eFlex, although the eFlex should still be secured to the shelf in some manner and the shelf must be rated for the load.

Let’s end the discussion by talking about some balance of system material for battery combining. Almost all eFlex applications will require at least 2 eFlex, and so combining eFlex together is a necessary task.

Daisy chaining eFlex power cables is not recommended. The result is uneven cable lengths between the battery and the inverter – these differences can result in shorter battery life or even trigger the battery management system shutdown. So the last step in eFlex design is how to combine the eFlex in a cost-effective manner.

There are multiple ways to do this, but a metal gutter fits well between the eFlex and the inverter, and if it is large enough, such as an 8” x 8” square tube gutter, all the material necessary for battery combining can be installed inside.

This device is called a power distribution block. It is UL-listed, commonly rated for 75C. To find the right sized power distribution block, one must identify the size of the inverter cable, DC amperage, and the size of the battery cable. It must also be listed for fine stranded wire.

Most power distribution blocks require ferrules with fine stranded wire.  Ferrules are larger than the wire sizes which they contain – a ferrule for #1 wire has a diameter about equal to 3/0 wire.  After additional research, a power distribution block rated for fine stranded wires is found, and it is a simple matter of seeing if the inverter cable and battery cables are appropriately sized. Alternately, a power distribution block with studs for ring terminals could be used.

Let’s digress for a moment to discuss battery cable. There is a UL listing for “battery cable” but it is not available in wide supply. The product literature of the power distribution block mentions DLO cable, which is a fine stranded wire which is similarly robust like battery cable, is more widely available, and is also UL-listed.

Note that that finely stranded wire has a larger diameter than traditional stranded wire, and it can flare out when the jacket is stripped off. So when using a #1 AWG cable with a power distribution block, it is important to find one with a 2/0 terminal so that the finely stranded wire can land on the terminal port with ease.

We also need to determine the cable size for the inverter. This popular battery inverter mentions a maximum inverter charging capability of 185A, as well as a 9kW rating. 9kW divided by a nominal voltage is 51.2V is 175 A. Using the greater of the two, with an additional 25% safety factor capacity, results in a 218A necessary cable rating. This results in 3/0 cable using 75C terminals, although 4/0 is often used due to better availability.

At any rate, the power distribution block is rated for 500 kcmil wire and so will accept a 4/0 cable readily.

The end result is that we need a short run of 4/0 DLO cable, this power distribution block, an 8” metal gutter, another run of 2/0 DLO cable, some strain relief connectors, and some compression ring terminals. The overall balance of system cost is kept to a minimum while providing a quality, code compliant, and fine looking installation.

As a final note, there are power distribution blocks available with studs rather than terminal ports. In general, the studs are rated for 90C, allowing amperage calculations to be performed at the higher temperature rating. This can reduce the required cable size.

That is all the time we have for today. Remember a copy of this presentation is available in the dealer resource section of our website, and installers, distributors, and designers may always email with project questions at any point in the project.

Lithium Battery Systems


Fortress Power manufactures lithium-ion phosphate batteries and I have the pleasure to work in their engineering department in field applications and product management.  We are headquartered in Southampton, PA, just north of Philadelphia.  The office roof has multiple solar arrays directly wired into our testing lab, where we test our batteries on various inverters systems.

We do testing, research and development, sales, technical support, and have coastal logistics to expedite shipments directly to material distributors throughout the country. We provide technical support to installers all over the world. We’ve shipped over 100 MWh of lithium iron phosphate batteries, for behind the meter applications and also directly to city and utility services, including transit.

We will start with an industry overview, talk a little about lithium chemistry, and then get into basic design principles that homeowners, designers, and installers should be aware of heading into a project to ease installation woes. We will also develop a better understanding of solar and battery systems in general.

 So with that, let’s dive into our class material.

This is an exciting time in the rapidly changing battery industry. Fifteen years ago all solar projects had batteries, but for the last decade, the grid-tied solar market largely went without batteries. But our NABCEP webinar audience indicates 35% of their solar projects now have batteries – its definitely on an upward trend.

Many of us with battery experience started in off-grid applications with lead acid. If you are budget constrained and currently living off grid with lead acid, then I suggest you stick with your existing battery chemistry to avoid surprises and instead upgrade the entire system down the road, after your charge controller failure. Older charge controllers simply lack the finesse required for the slightly different voltage settings of a lithium system. Likewise, off gridders with experience living on a small amount of lead acid are likely much harder on their batteries than they realize. To get a battery warranted for ten years, it is necessary to put on some protections which might surprise those who feel they already know everything there is to know about battery systems.

This is a new industry with new applications. For almost all professionally installed projects, lithium batteries are the solution for today. In fact, the market leading lithium battery chemistry (lithium cobalt) is being replaced with the next generation chemistry – lithium iron phosphate.

So lead acid still has a role to play when trying to maximize the life of charge controllers and battey inverters that are apporaching a decade old and designed with lead acid in mind, but that type of application really isn’t where most battery applications are today. For batteries that are grid connected, or for professionally designed and installed off grid applications, lithium is the clear leader and that will be the focus of this class.

For solar installers without battery experience, the shift from simple batteryless grid-tie solar with net-metering to battery systems optimized for zero outflow and variable rate structures can be overwhelming, but I would suggest it is quite fun, because like the solar industry a decade ago, system designers now have to improve their design know-how and understand at a deeper level the fundamentals of the systems being installed.

Growth rates in the battery industry is exciting, like the solar industry ten years ago. Even during the coronavirus, the US residential battery market has grown! There is  robust growth in the utility and commercial sector, but also in the residential market. Every solar installer should be quoting battery systems to their customers – a solar array without a battery is an incomplete system. Solar battery customers are more tolerant of a longer payback period, and sometimes the battery can improve the system economics – especially in markets which lack good consumer-owned solar policy.

Lithium Battery Chemistry

This is a nail test demonstrating the difference between a lithium cobalt pouch battery and a lithium iron phosphate pouch battery, so this is an apples-to-apples comparison of lithium pouch batteries with different lithium chemistries. The footage is quite dramatic here, reminiscent of airplane fire disasters in recent years.

The lithium iron phosphate battery results are quite different. It’s the exact same test. This test part of a battery listing, as it simulates what can happen to a battery during an internal fault. The fault is simulated by drilling straight through the battery. But lithium iron phosphate has a higher combustion temperature than lithium cobalt – the difference is only the chemistry! So whereas in the lithium cobalt battery there is an explosion, with lithium iron phosphate the nail can be driven straight through.

So now you know what a nail test demonstrates, as by the end of class you will learn how to properly evaluate grid-tied battery system design.

Batteries are expensive, but only recently have they become good enough to reduce an electric bill. Those economics boil down to cycle cost if the battery is used daily – and not all batteries are used daily. Those other batteries might prioritize low upfront cost (such as lead acid) or faster discharge rates (such as lithium cobalt), but ultimately lithium iron phosphate has the lowest overall cycle cost, and that is important for applications where there battery is used to reduce the electric bill. .

As a mechanical engineer, I kind of always hated chemistry and biology and pretty much anything organic except for a few humanities classes. But with batteries, the specifications all trace back to chemistry. We only have so much time in class, so we might have to get to the actual chemistry part later, but compared to lead acid, one pragmatic advantage of lithium is that is lightweight compared to lead acid. The cost advantages are also important, but for the same amount of storage or space, that weight is important. At the same time, lithium batteries are still heavy. It can make sense to have one large heavy battery for project simplicity, and on a different jobsite it can make sense to have multiple smaller batteries, at a slightly higher cost, so that the batteries can be carried up stairs easily or mounted on walls. So even though lithium can hold more power per pound than lead acid, perhaps it is better to state they have better storage density than lead acid but remember that they are still quite heavy.

These are sealed batteries. In code, the ventilation requirements of unsealed batteries are explicitly not applied to sealed batteries. At the same time, it is easy to imagine future code may consider emergency ventilation requirements given those nail tests mentioned previously, such as when a battery fails when it is not supposed to. LG Chem recently recalled 5% of its lithium cobalt residential batteries, after a likely thermal runaway event on a utility project site in Arizona blew up the shipping container the batteries were stored in. As you learn more about the nuance between lithium cobalt and lithium iron phosphate batteries, you will become frustrated that National Electric Code differentiates between all types of battery chemistries except lithium cobalt vs. lithium iron phosphate, even though they operate at different native cell voltages. But it is not yet time to complicate our battery knowledge. This is a two hour introductory class, and if you are already an expert and would prefer to skip ahead, then do so. But we’re going to try to keep things introductory for this class.

There’s a zero maintenance philosophy behind lithium batteries. If you are commissioning a lithium battery system, particularly on a retrofit using legacy products designed for lead acid voltages, you might have some post commissioning work to do if you do not confirm that the batteries are indeed charging at the correct voltage levels.

When I first started in solar, I helped build a mobile solar trailer which powered a radio promotional broadcast in a parking lot. There was a bouncy castle for kids to jump in, that was powered by a loud gas generator, and we suggested that the generator should be turned off and the bouncy castle run off our solar trailer. An hour later intro the broadcast, there were children screaming inside the collapsing bouncy castle, and it definitely felt like solar power had become quite suspicious, but actually a balloon had been sucked into the bouncy castle’s air intake and it really wasn’t the battery’s fault or even the solar array’s fault for that matter. The moral of the story is that a technology when properly operated can be zero maintenance and that there is also a need for monitoring, and also that things can go wrong despite the best technical support but very rarely is it actually the battery’s fault.

So the additional technology built into professional lithium battery packs needs to be properly confirmed during commissioning, but if you want a battery where you can set it and forget it, but use it daily over a long period of time, including fast charge and discharge rates associated with storing less than a few days worth of storage capacity, then you can assume lithium is a better choice than lead acid.

When it comes to lithium chemistry, lithium cobalt has a higher native voltage than lithium iron phosphate. That has numerous implications with regard to performance, such as combustion temperature as we previously saw, but also discharge rates and longevity. But specifically related to voltage, one NEC quirk is that systems operating under 60 DC volts are subject to fewer requirements than systems operating over 60 volts.

The battery industry as well as NEC will still call these batteries 48 volts, and nearly all quality 48V battery inverters are compatible with 48V lithium batteries. But in reality a 48V nominal battery has an operating voltage that is higher than 48V, regardless of chemistry.  The nominal voltage is closer to the battery capacity at a 20% state of charge. In other words, the 48V classification is a broad definition, like categorizing both a Honda and a Tesla a mid-size sedan without any further detail.

Comparing the cell voltage of lithium cobalt versus iron phosphate, the true nominal voltage is lower for iron phosphate than cobalt, which means a fully charged 48 volt lithium iron phosphate battery will have an operating range at just under 60 volts, whereas lithium cobalt battery with the same number of cells (as is common) will operate at a range just over that 60V NEC benchmark, which triggers more design requirements.

So it is common to find lithium cobalt manufacturers building high voltage batteries and repacking them as high voltage residential batteries, but lithium iron phosphate is better optimized for lower voltages, which are safer in general and also with specific respect to fire, but also with respect to cell degradation and internal combustion temperature.

Here is the warranty of one of the more popular lithium cobalt systems on the market. Product warranties are a great source of information, and for an investment of this size, it is certainly worth reading the manufacturer warranty – the maximum operating range of a battery is not its ideal operating range and the same manufacturer warranty can differ based on operating conditions.

This warranty is for a 14 kilowatt hour lithium cobalt battery bank, and we see that it is a warranty based on total throughput rather than cycles.  So this particular battery has a 37,800 kilowatt-hour or 10-year warranty, whichever comes first.

If you were cycling that battery at a 80 depth of discharge each day, the math shows the warranty has an equivalent of 3375 cycles, which divided by 365 days per year is slightly less than a 10 year warranty. While this battery may not be used every day and thereby obtain the ten year warranty, it is reasonable to assume a battery with a 10 year warranty and more robust throughput warranty will have a substantially longer product life and make it through the duration of the manufacturer warranty term. More on that later.

Here is the an alternate 18.5 kilowatt hour iron phosphate battery, rated for 87,600 kilowatt hours of throughput for a duration of 10 years. If cycled once per day at 80% depth of discharge, does that mean it will last 16 years? Or if the battery is never used, does that mean it will last forever? The answer is no in both cases, although a lithium battery that is regularly used will in fact last longer than one that is rarely used, because of how that chemistry actually degrades.

So can you read a warranty to determine which battery is better? Ultimately a long warranty comes down to using a properly sized battery bank at the optimal battery settings, rather than the maximum battery settings, no matter what a salesman actually tells you. A battery that is used in a more aggressive manner will not last as long as a battery used within a more narrow operating range. There is no such thing as a “100% depth of discharge battery” verses an “80% depth of discharge battery” nor is there a faster lithium battery verses a slower lithium battery. The chemistry drives the battery performance and the warranty and ultimately the product life is driven by the operating conditions and battery management system. In a lab at ideal conditions, a cheap lithium ion battery will perform very close to a premium lithium ion battery of the same chemistry, but as a package in the field over a warranted period of time longer than the industry itself, the final product requires more than just the cells and volts to be successful.

Cycle ratings, for that matter, are a bit of a marketing gimmick; the real focus should be on throughput warranties and warranted life. This is because cycles are very hard to track, and the consumer does not have access to the manufacturer real cycle data, because like an onboard computer on a vehicle, a battery computer can become dangerous if the consumer can hack the data logs and safety constraints.

In a perfect world, a battery would have one cycle per day, charging during the day and discharging at night. A perfect charge and discharge cycle would be great and will result in a longer lasting battery when right sized.  But we all know how we charge and discharge our cell phone batteries: intermittently. And the same is true of building energy use. A solar battery may charge and discharge once per day, or more or less than that, depending on a variety of conditions.

The point is that a daily use lithium cobalt battery warranty will not last ten years under full use, but a lithium iron phosphate will exceed that 10 year warranty term. You can find lithium iron phosphate batteries with unlimited partial cycle ratings, or cycle ratings that project a 30 year life, but almost always those battery warranties are constrained by operating conditions which are less than the battery maximum rating, as well as less than the battery maximum temperature rating, and then also constrained by an annual year warranty. A ten year warranty term is as good as it gets, and the projected lifetime should not exceed more than 50% beyond that term.

An intermittent power supply like solar, as well as variable loads, means the battery is going to be charged and discharged and charged and discharged.  There are partly cloudy days which can have multiple charge and discharge cycles. When the cycle rating is higher than the warranty annual term, the battery owner can be confident that the battery life will expire last the duration of the warranty and much longer, so long as the battery is used at its recommended parameters. And there are applications where a reduced warranty at expanded parameters is justifiable – and the actual product life would be about the same if those expanded parameters are relatively infrequent controlled as opposed to occurring on a regulars basis.

This is a simulation using commercial EnergyToolbase software, and the graphic reflects analysis we do for battery right-sizing. In green is the solar production and yellow is the battery production. In light blue, we see the solar production coming up and then it’s partly cloudy and the solar production is going back down and then back up again and coincidently, the battery  is charging, and then discharges when its cloudy, and then it charges again when the sun comes back until the sun sets. The takeaway here is that cycles are frequent, incomplete, and difficult to track, but regardless, chemistry is chemistry and lithium iron phosphate is going to last much longer than lithium cobalt.

Before leaving this slide, what we see modeled here is commercial facility electrical demand remaining perfectly level. In a “time of use” scenario where daytime electricity is billed at a higher rate, all of the power during this time would come from the battery. Instead, this battery is only managing the building demand, keeping the total building power below a defined threshold. Peak demand is often the commercial facility’s most expensive electricity and it can be targeted with a relatively small battery that only needs to run for a couple hours per day, if that.  There’s a lot of different ways you can use lithium batteries in a cost-effective manner.

 In continuing to answer the question, “Why would you use lithium iron phosphate over lithium cobalt the takeaway is that lithium iron phosphate compared to lithium cobalt has increased fire safety and a longer life.  When you combine the two, that’s really what customers want. But, in addition to that longer life, it also translates into a lower total cycle cost.  Where maybe the higher voltages that lend to slightly faster charge and discharge might be better for an electric vehicle application, when it comes to integrating batteries into buildings you want the battery to last as long as possible.

You want it to be a daily use battery where you’re competing with the cost per kilowatt hour not necessarily the cost of gasoline!  You need that economic competition and advantage of the longer cycles, better throughput, total longer life of the battery is the advantage of lithium iron phosphate over lithium cobalt.

Market Applications

Let’s talk about some market applications of batteries.  This is a slide for Georgia Power’s rate structures and I would always encourage solar installers to get on the websites of their local utilities and look for variable rate structures.  They might come in the form of a free nights and weekends plan,  or a “time of use” plan that will adjust your electric rate

based on time of day.  

Georgia Power has a smart usage plan that basically takes a flat energy rate and converts it into a demand charge for them. Georgia has a very limited net metering program right now, but historically, they have not had net metering.  How do you install a solar array on a power grid that doesn’t have a net metering policy in place?  Start to look at the variable rate structures and design an entire system around that.

Here at Fortress Power, we can assure you, a solar array without a battery is an incomplete system!  All joking aside, it is!  When you combine an array with a battery, more risk is mitigated and you can use variable rate structures to save even more money. For example, here is a $30,000 solar project in the southeast US.  Say it’s a larger home with an electric vehicle on site.  They’re using 2,500 kilowatt hours of electricity per month at $0.12/kilowatt hour and a variable rate structure.

A variable rate structure is something you’re probably already familiar with, but if you aren’t I will explain.  Often described as, “doing your laundry at night”, this structure charges more at peak times and less at other times like nights and weekends. Normally you would have a flat fixed rate for your electricity, whereas demand charges are more common to commercial businesses.  The demand charge availability for residential in this example is $0.10 per kilowatt then they give you an off peak electric rate of $0.01/kwh. The flat rate is $0.12/kwh.  If you work non-traditional hours as so many do now, this can be a real savings. You aren’t in the shower at peak rate hours, etc The household must evaluate usage and decide.

Disclaimer: This is a sample budget. Obviously, this is an educational example. We’re not talking about the actual price; instead we’re talking more loosely, so the actual price and availability of the actual product is subject to change.  Now, back to the show.

Let’s say we add this 8 kilowatt battery inverter. We can reduce the power draw of the home.  Let’s say their old power draw was 15 kilowatts and now it is 7 kilowatts. They didn’t have the demand charge on their old bill, but now they have the demand charge. We’re assuming the solar array provides enough energy to charge the battery.  These are conservative assumptions. Run the PVWatts model for more detail, but we’re assuming the battery and the solar array can fully offset the customer’s peak energy and then what they have left is off-peak energy.

So you can see that even though the demand charge is $70/month the access to that cheap electricity means they have quite the savings on their electric bill so we took at that this 30 project budget.  We apply the federal tax credit and looking at $200 of monthly savings.  Your $24,000 budget after tax credit divided by $20,000 a year in cost savings means you get a 10-year payback.

While a 10-year payback may not initially dazzle you, think about it: First, all of your products involved have a 10-year warranty. Add the energy security of having that solar array and battery on site.  Now, when a hurricane takes out the power grid, you know your family still has power.  The neighbors won’t.   They are throwing out food, but you aren’t, and that is a massive savings itself. Medicines are the proper temperature.  Your kids can still distance learn. You can still work remotely.

You can add an electric vehicle, something that is going to be more and more prevalent over time. But batteries aren’t just for electric cars.

With the price of lithium iron phosphate coming down, new markets emerge for solar as well.  Now, thanks to affordable batteries, you can have an array in an area where there’s no net metering policy at all. Prior to the battery, solar payback would be non-existent.  All of a sudden, you add a battery and a variable rate structure, and any consumer can get an array if they wish. Maybe the utility does not buy back the solar at a good rate; just add a battery!

Another very popular application is commercial demand management. This is where even just a small solar array that doesn’t do 100% of the building’s electric bill and a small battery that doesn’t do 100% of the building’s demand,  target exactly where you have enough solar to charge the battery and enough battery capacity to dispatch the battery exactly when it’s needed on a demand charge.

The demand component is your instantaneous power draw.  That’s the most expensive component of the bill, so even a small solar array and a small battery on a commercial project can result in a rather large savings, especially if your business has a load profile where you have a energy intense process for a very short period of time, and the rest of the day you’re operating at far below that demand, you can use your commercial demand management application and have it be quite cost effective in that scenario.

Now, we have batteries applications for South Eastern PA Transit Authority, or SEPTA, here in Philadelphia is using our batteries for their applications. Telecom; there’s a lot of small applications for this technology as well and so when you’re up against a variable rate structure where it’s worthwhile to use batteries well

Here’s some battery budget estimates. These aren’t necessarily what we’re selling the batteries for but they’re ballpark figures and when you take the throughput warranty of the batteries and you divide by their capacity you’re seeing prices of $0.11-0.14/kilowatt hour, but don’t forget that the batteries do apply for the federal tax credit so when you add a solar project to it that drops the price to $0.08- 0.10/kilowatt hour.

Now, explaining some of these differences, here we basically have a larger battery and a smaller battery, and it shouldn’t surprise anyone that the larger battery has a little bit lower cost at the end of the useful life than the smaller battery. So, just some ballpark figures, what I would say is if you look at your time of use rate structures and there’s a differential of  $0.08-0.10/kilowatt hour, that’s an application where you can start putting in a battery on site and then particularly when the time of use rate structures are even larger than that, then you start to really make some money.  

Here’s an electrical cooperative in Mississippi. Are you shocked that Mississippi does not have good solar policy? They have very bad solar policy, at least from the homeowner perspective, but even when you have bad solar policy you can still have a great variable rate structure. And that’s all a solar battery really needs.  Here we see Coast Electric’s variable rate structure.  Normally they’re paying around $0.09/kilowatt hour for their electricity for three hours a day they increase it by four times to $0.36 and the rest of the year they cut it in half down to $0.05 and so you see this differential that’s greater than $0.08- 0.10. That means this is a great time of use rate structure for solar + battery.

Here’s a study by National Renewable Energy Labs of demand charges throughout the country. What I like about this graph is we can see there are high demand charges in California, sure, but there are also high demand charges in the middle of the United States, which is less obvious.  You can find, especially in rural areas, the demand charges start to uptick and so you might find yourself putting a battery and maybe a small solar array, just enough to charge the battery, just enough to lower the building’s peak demand.

That’s another very robust, viable application of solar power and it’s really the next step in what we need for solar to power 100% of the grid or solar and batteries, you know, to really transform our country into more than just for the luxury energy market, but for the entire United States. Our next step in growth is adding batteries back into solar projects. So here are some project photos. It’s all about having that backup power.

Design Principles

Now we’re going to cover some design principles. These are things that you really need to know before you begin your solar project with a battery bank.  These are our sanity checks, to make sure you’re on the right track.  

Lithium works best at a two hour discharge rate.  One of our competitors in the lithium cobalt industry packages an inverter along with the battery and if you look at the inverter size and the battery bank size you’ll realize it’s impossible to discharge that battery faster than two hours with that very popular package.

The same thing applies with lithium iron phosphate.  You know, both technologies, lithium cobalt and lithium-ion phosphate, have a longer life at a slower discharge rate and you’ll see that reflected into warranty issues. So, faster rates of discharge are possible but they’re typically only reserved for commercial application like demand management. The fact of the matter is if you discharge your battery over one hour time frame that’s not going to give you a lot of backup power capability and so that means that faster rates of discharge are really reserved for larger commercial applications and not something that is particularly useful to the residential market.

So what does that mean? If I say lithium is a two hour battery, this is kind of a call back to lithium versus lead acid, and so what will happen when you discharge a lithium battery at a two hour rate is that it’s its voltage will remain relatively flat until the battery is mostly drained. If you go and apply that same load to an equivalent size lead acid battery. It’s curve is going to be way further down.  It might only have half of its useful life or even less.

The voltage might drop so much that the load might even not turn on, and you can see that reflected in lead acid spec sheets. So, this is a solar lead acid spec sheet. They don’t give you a two hour discharge rate on a lead acid battery. The spec sheet says they’re rated for a 10-hour discharge rate.

If you fully discharge the battery at anything less than that, the battery is not going to work!  You’re going to get into reliability issues. The customer may turn on heavy loads and the voltages will drop so much that those loads will turn off and now all of a sudden you’re doing the opposite of what you wanted the battery bank to do: it’s turning off your electricity rather than keeping it on! So, particularly when we’re looking at smaller battery banks, lithium-ion is a clear winner.  Lead acid is just not rated for rapid discharge, just buying a few kilowatt hours, but not running a relatively high load off of it.

Now all that said, there still is a limit to lithium technology. We’re not giving it a 30 minute discharge rate, we’re giving it a two-hour discharge rate, and maybe a one-hour discharge rate with fewer warranted cycles in a commercial specific design scenario. Lithium-ion is still a two-hour battery and so for a quick sanity check to see if the design has an undersized battery, take the battery capacity and divide by the AC inverter output, and if you’re over a two hours discharge, you’re in the clear on your design.  You’ve cleared the minimum battery bank size that’s necessary in order to pass this this test.

The mistake that we see in the field from our battery designers is they’ll just buy the smallest 48-volt battery and the largest 48-volt battery inverter.

That results in a 5 kilowatt battery on an eight kilowatt inverter. If I ran this inverter at full capacity it would only take me theoretically 40 minutes to drain that battery. That’s not within the specifications of the battery. So, you’re going to run just like you would with lead acid at less than a 10-hour rate or less than a 6-hour rate. You’re going to start running into issues with lithium at less than a two hour rate unless it’s a special circumstance.

Another design mistake we see is designers sometimes do not respect the maximum inverter output on a battery inverter! If you put more load than the battery inverter can handle, it’s going to turn off.  It’s going to enact its own safety precautions. If you try to overload a battery inverter, you can insert reliability issues into the equation. We need a better understanding of how battery inverters actually work.  A solar inverter is one way. If you get rid of the batteries, it just takes all the power from the solar array and dumps it into the grid.  The grid does not back feed the solar inverter and charge up the solar array.

In other words, a battery inverter is more complicated than a solar inverter. It’s more expensive.

There’s actually two sides to a battery inverter as well. There’s the grid side and the backup side, which creates its own signal. The grid side synchronizes to the signal but guess what?  If the grid goes out, the synchronousness side goes out with it. The asynchronous side remains running, but what this means is that the load on the asynchronous backup side, regardless of the mode of operation of the inverter, regardless if the grid’s up or the grids down, you can’t overload this backup side of the inverter.  Otherwise it’s going to lose power and drop power.

 That actually is a National Electric Code violation, so National Electric Code says if you’re installing a standby source of power, your power supply has to be able to power the load that it’s connected to, especially in a backup circumstance or an emergency circumstance, because the customer is relying on that power.

So where you can run afoul of your design is looking at your usage data and if you’re fortunate enough to have a utility account with a smart meter, you can go and see the 15 minute or 30 minute interval data.

The interval data is telling me that this is not going to exceed 14 kilowatts of power, and so you can put that on a 16 kilowatt battery inverter and be just fine, even have a little bit of headroom.

The reason why that can result in design errors is that loads are not spiky. Loads are not 15-minute spikes, they could be 5 minute spikes and if that 5-minute spike is averaged over a 15-minute interval on that 15-minute interval data it can show up as a much smaller load than what it actually is.  Install a consumption monitor on your electric service panel and get one minute data or five second data, and you’ll see a much different picture.

The fact is, it gets really complicated, and if you read too far into it and if you try and push the limits of your design too far, you’re going to run into mistakes. A big example is that battery inverters have what they call surge modes.  Here’s a 10 second surge mode, here’s a 0.1 second surge mode.  You can see this battery inverter can surge. It’s a 8 kilowatt inverter but it can surge to 16-25 kilowatts. We’re talking about a matter of seconds, not a matter of minutes. The fact is if you have an 8 kilowatt battery inverter and it runs for more than one minute at above eight kilowatts it’s going to turn off.

 At that point, everything on the backup side, everything on the asynchronous side of the inverter, is going to turn off with it, and those are, in fact, your critical loads.  Those are the loads that you don’t want turning off, so what is this burst mode is really referring to, and the thing is, there are certain loads that have very quick charge and discharge, one of those loads is the inverter systems.

The charge controllers themselves, because they have to be capable of supplying this very rapid burst, (and that means they’re also capable of sucking in that much electricity very rapidly) so when you turn on the inverter for the first time these capacitors suck up all the energy, they’re connected too and that can that can result in that burst or surge rating. So, if you really wade into the spec sheets.

You can find the burst mode for the battery and the burst mode for the inverter and do your sizing from that but as long as you stick to this kind of “down and dirty” two-hour sizing principle, you’re going to be in the clear.

Here’s another quirk of solar battery design. There’s not much of a cost difference between a four kilowatt battery inverter and an eight kilowatt battery inverter.  So, we see a lot of solar installers opting for the 8-kilowatt battery inverter.

What you need to remember is that when you increase your battery size, say, you double your battery size, you’re also going to increase the minimum battery size required to go along with that battery. Not all off-grid customers go for the minimum battery size, but if you’re only planning to have your solar array and battery there for two-hour long power outages, or even longer if it’s during the day, you don’t want your battery bank to be too small. If you take the inverse size of that battery inverter and multiply it by two, running that battery inverter for two hours will to give you the minimum kilowatt-hour rating of the battery bank to have good reliable grid electricity.

Is 8 kilowatts of battery inverter, which is a very popular battery inverter size, enough to power a whole house?  The answer is no. Eight kilowatts is not enough to run air conditioning units. People think having an energy efficient heat pump that only runs at 3 kilowatts, and it doesn’t run at10 kilowatts, I know it’s on a 50 amp breaker but it’s only running at 20 amps, so they’ll never get near the maximum breaker capacity.

Sadly, there’s no such thing as a free lunch.  If an electrician put that air conditioning unit on a 50 amp breaker there’s a reason for it, probably because the manufacturer recommended it. Breaker size heat pumps and air conditioning units, even if they’re very efficient they can be very inefficient for a very short period of time.

This isn’t about the operation of the heat pump itself, but some auxiliary circuits, such as turning on the defrost inside the unit. So, if your unit’s working too hard and it ices up, it has some electric heat tape inside of it that turns on to rapidly fall out the system so that it you can run it really hard again. These heat pumps can use the same electric tape to provide a short burst of electric heat in the winter. If you want instant heat, the heat pump can help out with that, and then it’ll transition a more energy efficient mode of operation.

Most heat pumps have 10 kilowatts of heat tape inside them and so one air conditioner unit even if it’s only running at 3-4 kilowatts under normal operation is still too large for 8 kilowatts battery inverters. You’re not going to have your air conditioners on it. Prioritize other loads, which we’ll get into in a minute.  So, with 8 kilowatts, think about the loads that you absolutely need when the power goes out. Medical equipment, main bedroom, your entry and exit to the house, internet and television and you don’t want the the food in the fridge to go bad.

True critical loads are the most basic functions of a home, without air conditioning or hot water heating.  But people want more than that.  The air conditioner is going to be something that the customer wants to have on during a blackout.  For most large homes in the US, you’re going to need more than 10 kilowatts of battery inverter.

What if you use two battery inverters that are 8 kilowatts?  The critical loads we just talked about run on one eight kilowatt inverter, but the second inverter can”t always handle the air conditioning and hot water. A dozen days a year your air conditioner unit runs a 10, and so if you’re not controlling these loads and they all turn on at the same time, that is well above an eight kilowatt inverter, so an unmanaged load you really have to choose between hot water or air conditioning.  You can’t have both at the same time.  If you want both, you need three 8 kilowatt battery inverters.

Remember: every single time you increase the size of your battery inverter you’re also going to increase that minimum size of your battery bank.

There are load management devices that you can implement to make sure your hot water tank doesn’t run at the same time as your air conditioner.  In fact, I love electric tank water heaters on solar battery systems because the tank itself can use be used to store the hot water.

So, if you have a load management system that can turn on your hot water tank during off-peak times and then turn it off when the air conditioner is running, then you might be able to squeeze more onto that backup load panel.  Just keep in mind the solution that works has to work when it’s grid tied as well as when it’s off grid! The minute that backup load panel goes above the inverter output rank rating, the system will turn off.

Smart home and smart appliances are exciting from an energy management perspective.  An electric tank water heater uses about 40 amps of power and you can find 40 amp programmable smart controllers to put in line to turn those tanks on and off.  Really, all you need to do is talk to your home consumption so that during your electric tank and then you can turn it back on again during off peak.  Solutions are available on the market. The real question is reliability.

Lumen makes a energy manager or basically they have a lot of disconnects. They wire them to all your major circuits right next to your electric service panel so they can turn them on and off based on the power usage of those individual circuits. Installers generally operate on the KISS principle and want the smallest possible configuration andthats understandable.

Obviously, in certain markets, you don’t need to have air conditioner on a backup panel, so location makes a big difference. A Texan would rather have all the food in the fridge spoil than to go without air conditioning, but in Maine, not so much.  You have to compare the implementation cost of your energy management system as well as its reliability. We aren’t trying to do a commercial for Lumen but a hardwired system like theirs is going to be more reliable than a wireless system.  Trying to hodgepodge together a variety of smart home devices is tricky and might cost more.  In the end, the client has to decide.

As a final note, and maybe we’ll talk about this more in the next class, you couldn’t put an undersized battery that violates that two hour rule because this is 18.5 kilowatt hours on 16 kilowatts of inverter capacity. You would put all of these devices onto two 8 kilowatt inverters so that during grid tied operation you have that reliability.

Another design rule:  Don’t neglect generators!  Generators can go very far to reduce your minimum battery bank sizing.  

Here is a graph where we’ve taken the month to month and divided it into day-to-day energy use of the building. Every day the building uses power, every day the solar array charges the battery, so we’ve just taken the building consumption, using 12-month electric bills, taken the PVWatts solar production figures, and we’ve charted it out for every day of the year.

Assuming we’re not running a generator, we need to make sure we don’t run out of electricity, and so our minimum battery bank requirement is huge. If we have a generator turn on to cover this interval, this is only for extreme weather systems like a summer hurricane, with 4-5 days of cloudy conditions, you can either buy 4-5 days worth of batteries or you can buy a little bit more than one day’s worth of batteries for off-grid, and let the generator take over.

In the past, the solution for off-grid without a generator was to buy four days worth of energy storage to avoid generator usage.  Now installers have to decide between four days worth of lead acid or one day’s worth of lithium iron phosphate.

Keep in mind that you really can’t eliminate the generator entirely in an off-grid scenario, and many customers are designing for the minimum battery size for protection during short power outages.  During the day, save the generator for outages that are longer than two hours. The size of the battery bank when you’re grid tied in particular shrinks down even further.

The last thing I want to talk about are battery management systems. They do more than just monitor voltage!

A quality battery management system will disconnect the battery during failure state modes like overcharge, discharge, extreme temperatures, or when those inrush currents are too high.  

Battery management systems for home hobbyists are weaker than battery management systems for a whole house. For small appliances, they’re only using a couple of amps like a server. A whole house 48 volt battery on the DC side can get up to well over 100 amps. The traditional way to manage your energy and a battery management system is to use these what’s called a MOSFET, which stands for Metal Oxide Silicon Field Effect Transistor or Metal Oxide Semiconductor Field Effect Transistor. This is also called as IGFET meaning Insulated Gate Field Effect Transistor, and they’re great for low amperage solutions but they do not withstand the high current that’s required for nearing whole house backup!  Battery management systems use embedded processors instead on their own circuit boards and are designed to handle that high current.

Lastly, a battery management system can communicate directly with an inverter to manage charge settings. Most installers think this means you do not have to program inverter charge settings – this is not true. Settings still need to be programmed, because the inverter needs a fall back in case communication fails, and also some settings are site specific, such as if the homeowner wants to use the battery to its greatest extent or reserve some capacity for an unplanned power outage.

The greatest advantage of closed loop communication is being able to undersize the battery bank, such as combining battery combinations which may in fact violate the 2 hour sizing principle we discussed at the top of the hour. So, closed loop communication is most advantageous when the project has a limited budget, and the battery is not relied upon for daily use. In an offgrid system, the battery banks need to be larger and the precise controls that come with closed loop communication are less advantageous.

Well, thank you all for coming, we’re out of time today but for next time, we’re going to discuss installation pain points and I’ll talk more more about battery management systems and their features and functionalities.  If you have any questions, ask them on the survey and I will be happy to answer them.  I hope you will come join me in the next video! At Fortress Power, we know how valuable your time is and we sincerely appreciate you taking time to come learn with us.

[button outline=”true” url=””] Part Two [/button]