Inspector Wants A Load Calc For Battery Backup

[pv_n00b]

I feel your pain. In a meeting with an inspector and his (non technical) boss a while back where I had challenged the inspector on a point of code, I was right and he was wrong. When I showed him the code language that was clearly in opposition to his position, he just fell back on, "I am the AHJ and I reserve the right to interpret or revise the NEC any way I see fit." His boss just looked at me and shrugged. End of meeting.

Many in here will say that he can't do that. In an ideal world that would be true; in the real world, not so much.

There's the ideal world, the real world, and then there's Texas. Kind of like the right way, the wrong way, and the Navy way.

 

[WaveGuide]

For interconnected systems (such as the Tesla battery is, ... and even though it's called a "battery", which the inspector probably focus on, and we all know it's much more than that), I noticed the following in 706.16(F) which points to 710.15, which effectively says that the source when operating in island mode, is permitted to be less than the full load! (I'll let you read the details.)

But what this says, is that for an interactive system (i.e. one that operates in parallel .. at times with the primary source, and in island mode at other times) , that when it's operating in island mode that it is NOT required to meet the same sourcing requirements, or even all load requirements.

(It does have to supply the smallest non lighting load however.)

So this might be helpful in the next argument with an inspector on this point. BECAUSE the source is interactive, then it gets 710.15 when in island mode.

 

[Steve16]

If it's being used as an optional standby system (which most are for residential) it would fall under 702. If this is being used as an emergency power system in a hospital are you going to argue that the rules of 700 don't pertain?

The whole point is to avoid interruption of power. If the battery backup can't handle the load and either trips out the system or is constantly refreshing every X minutes without being able to transfer over what's the point?

 

[jaggedben]

...

The whole point is to avoid interruption of power.

That's inaccurate and overbroad.

First, standby systems are not UPSs, so it's not about momentary interruptions. Most standby systems take at least a second or two to switch over. Even those systems that can do it in a 10th of a may still drip your internet until it can reboot. I've yet to meet a client who is bothered by this. Those that care and understand what a UPS is typically already have a UPS or two. So that is not the point.

The point is to have power gor essential needs during long grid outages. Additionally, the point is to be able to use during outages the solar and battery systems that mainly serve other purposes. BTW that's unlike optional standby systems which only supply power to loads during outages. This is one element in arguing that 710 applies, and not 702, for PV/BESS systems that supply loads at other times.

If the battery backup can't handle the load and either trips out the system or is constantly refreshing every X minutes without being able to transfer over what's the point?

Tripping out the system can be a very effective poor man's load management. Think of things like laundry appliances and ovens, which won't turn back on by themselves. For a battery ESS that comes back on a couple seconds later this works rather well, as long as the overload isn't caused by something like an air conditioner that turns back on its own.

It could be argued that such a feature could be a way to comply with 702.4(B)(2)(b), although I admit it's a stretch.

Again, smooth transfer isn't necessarily the point and isn't necessary for these systems to serve a purpose.

...

For me the conclusion is twofold:
1) the disagreement between 702 and 710 needs to be cleared up.
2) 704.2(B)(2), requiring the system to serve the entire connected load or have load management, really just isn't justified and necessary for residential systems (no matter the type of backup power). These systems are optional, meaning the load served is also always optional. Seriously considering a PI to adjust this.

 

[tortuga]

For me the conclusion is twofold:
1) the disagreement between 702 and 710 needs to be cleared up.

I Agree, I think the key difference here is I have had to repair damaged generators (for example replace caps) after a severe overload, if the Enphase, Tesla or whatever battery unit is not damaged by a severe overload then I think its fine to skip 702 and go to 710. Generators with hi-tech overload protection as well.

 

[Steve16]

That's inaccurate and overbroad.

First, standby systems are not UPSs, so it's not about momentary interruptions. Most standby systems take at least a second or two to switch over. Even those systems that can do it in a 10th of a may still drip your internet until it can reboot. I've yet to meet a client who is bothered by this. Those that care and understand what a UPS is typically already have a UPS or two. So that is not the point.

The point is to have power gor essential needs during long grid outages. Additionally, the point is to be able to use during outages the solar and battery systems that mainly serve other purposes. BTW that's unlike optional standby systems which only supply power to loads during outages. This is one element in arguing that 710 applies, and not 702, for PV/BESS systems that supply loads at other times.



Tripping out the system can be a very effective poor man's load management. Think of things like laundry appliances and ovens, which won't turn back on by themselves. For a battery ESS that comes back on a couple seconds later this works rather well, as long as the overload isn't caused by something like an air conditioner that turns back on its own.

It could be argued that such a feature could be a way to comply with 702.4(B)(2)(b), although I admit it's a stretch.

Again, smooth transfer isn't necessarily the point and isn't necessary for these systems to serve a purpose.

...

For me the conclusion is twofold:
1) the disagreement between 702 and 710 needs to be cleared up.
2) 704.2(B)(2), requiring the system to serve the entire connected load or have load management, really just isn't justified and necessary for residential systems (no matter the type of backup power). These systems are optional, meaning the load served is also always optional. Seriously considering a PI to adjust this.

I should have worded my post better. I didn't mean it to sound like a ups.

But you refer to the fact that the point is to use with solar is not completely accurate. While that's a great convience if you have solar(and the most efficient way to use these), there are plenty of AC coupled systems that straight up do not need solar and can grid charge. They might not last more than a limited amount of time without the solar to recharge but its not required. They provide optional standby power and are using an automatic transfer switch which is specifically called out for under 702. If you want to say 710 applies in addition I don't think anyone would disagree.

Tripping out the system defeats the purpose of 704.2b2b. If the system resets itself and constantly tries to reboot I think you have a fair arguement but that would be manufacturer specific. But if it requires a physical reset of any kind there's no way it should be allowed.

The last couple generacs I installed have to have the inverter physically reset/rebooted if the draw from the house is too high. And guess what? It's nearly instantaneous transfer. None of the high load appliances (dryer, ac) stopped upon transfer unless the system was overloaded. It was so fast that the generac relays inside the transfer switch couldn't properly load shed the condenser units and I had to add their smart mods to make the system work.

Either way it's up to the AHJ and I know in my opinion it should be required

 

[WaveGuide]

... The whole point is to avoid interruption of power. ...

That might be your point, but I don't think it's the mission of the electrical code (and thus state laws) to keep the power on.

Rather it's the mission of the electrical code to make electricity safe.

Am I wrong about this? I'm open to listening to other points of view.

 

[zbang]

Rather it's the mission of the electrical code to make electricity safe.

Actually, isn't the original mission to avoid causing fires? Seems like most safety parts were added later.

 

[wwhitney]

Actually, isn't the original mission to avoid causing fires?

In that vein, what are the failure modes of an overloaded engine generator? Is starting a fire one of them?

Cheers, Wayne

 

[Steve16]

That might be your point, but I don't think it's the mission of the electrical code (and thus state laws) to keep the power on.

Rather it's the mission of the electrical code to make electricity safe.

Am I wrong about this? I'm open to listening to other points of view.

What if the homeowner is on some sort of life safety system that requires power?

Now that I say that, should it fall under emergency system and not optional?

 

[jaggedben]

I'm still mulling the best way to propose changes to 702.

Ideas (not mutally exclusive):
- Simply resolve conflict with 710 by finding a way to state that solar/battery systems and the like may comply with 710 and not 702.
- Propose less stringent requirements for one and two family dwellings
- Distinguish between systems that automatically recover from overload and those that don't, and state that for the former 702.4(B)(2)(a) only applies to connected load that turns back on automatically after power loss, or some such formula
- Distinguish minimum requirements for 'firm' sources (battery, generator) vs. total (also includes solar).
- clarify that 'connected load' means according to article 220
- propose a new section in 220 that would deal with off-grid situations other than article 700 and 701

 

[tortuga]

- Distinguish between systems that automatically recover from overload and those that don't, and state that for the former 702.4(B)(2)(a) only applies to connected load that turns back on automatically after power loss, or some such formula

I vote for something like that. I have seen generators get damaged from being overloaded. And the clueless homeowner repeatedly starting it again. . If the battery system or generator has a way to protect itself from starting into a overload then it should be exempt. If a battery system , generator or whatever source can be damaged from a overload then that section should apply.
There must be away for it to sense the impedance of the load its starting into and not start/restart if the load is too great?

 

[jaggedben]

I'm not familiar with how generators damage themselves. Don't they have internal circuit breakers? Or is the starting current the issue? Are you seeing this on generators with ATS or portable generators? I guess inverters have the advantage of having no inertia when voltage drops or current is too high.

I don't see how systems can reliably sense impedance before starting, especially for inductive loads or loads (air conditioner) that start automatically but not immediately after power is restored.

In a weird way, it's nice to know that some homeowners struggle with their generator systems too.

 

[tortuga]

I'm not familiar with how generators damage themselves. Don't they have internal circuit breakers? Or is the starting current the issue? Are you seeing this on generators with ATS or portable generators?

Generator with an ATS ~2015 model. And either the load changed or no one did a proper load calc.
The generator was a Generac 20 or 22kw, there is a voltage regulator with 3 large caps in it and it often blows one of those caps.

 

[mtnelect]

I am waiting for this summer in California on 100 plus degree days. When the grid goes down and all those residential PV get the life sucked out of them.

 

[pv_n00b]

It used to be that all PV was off-grid and designers had to spend a lot of time and have a high level of expertise to design these systems to work correctly. It was difficult to ensure that production and storage matched the load. Then grid-tied NEM came in and all a designer had to do was make sure the annual energy consumption from the utility bill was more than or equal to the annual PV energy production. The knowledge of how to design off grid systems was not past on except among a small group of designers still working in the off-grid sector.
Now we are coming full circle, batteries are coming back, both grid-tied and backup, and NEM is going away. Matching generation, storage, and load are now something designers have to do again and we have a lot of designers who have no idea how to do it.

 

[pv_n00b]

Since PV+BESS is more complicated to design there are software tools out there to help. I just got an email from Aurora Solar for their battery self-consumption tool that looks interesting. I'm sure there are others. If you are doing PV+BESS you are going to have to up your design game to keep up with the competition.

 

[WaveGuide]

IMatching generation, storage, and load are now something designers have to do again and we have a lot of designers who have no idea how to do it.

This is hard. I downloaded my past data for the last year from my electric company and put it in a database. Then took a cut of power used between 4pm and 8am the next day to find how big a battery I would need. Graphing this usage also helped, sometimes first sorting the data by usage so I could shoot for 80% coverage, and rely on the grid for heavier usage days. Also rewiring the sub panel to make it the critical loads panel was a challenge. What you need to think about is how the homeowner will limit their usage when the grid goes down. Smartload centers are starting to appear to help with this so the homeowner doesn't have to depower certain loads to meet battery supply capacity.

 

[ggunn]

It's hard to say where this is going. Like someone else said, when PV with battery systems were first being built, the users were mostly off-gridders who are well versed on the ins and outs of load management vs. battery capacity. With straight grid tied PV, it's pretty much set it and forget it, so customers by and large didn't know or care how it works. Enter PV with battery backup - PowerWalls et al. Customers walk into it thinking that it's going to run pretty much like straight grid tied PV and they could not be more wrong.

I know of a customer with a big house who spent a whole lot of money on a PV system with multiple PowerWalls. He called his PV company after that ice storm last February where his power was out for a while, and he was most upset because the first night of the outage his batteries were drained after six hours. He was threatening legal action, but the PV company looked at his monitoring and they could tell that he was running all his loads like the grid was up. They looked at his usage and his battery capacity, and (cipher, cipher, cipher...) - yep, about six hours. The system performed exactly as it was supposed to, and exactly as they had explained to him that it would. Turn off some loads or your batteries won't last the night.

Most people do not understand how electricity works. How do you accurately set a customer's expectations for a PV system with batteries when he doesn't know a Watt from a Volt or a kW from a kWh?

 

[pv_n00b]

So many people who get a PV+BESS will then go around telling everyone, I've got a backup system that will keep my whole house going for a week with no power.
Here's what the contractor said to them:
Your system will back up your house with minimal necessary loads but you have to turn off unnecessary loads or you will run your battery down in the first night.
What the homeowner heard:
Your system will back up your house blah blah blah. Hot tub party at my house when the power goes down, losers.

Realistically, some homeowners still think that having grid-tied PV provides backup power. People had to start putting it in large bold type in the contract that the PV system will not work when the utility is down, and even then that was just so they could be told where to find that text more easily when calling the contractor to complain about why their system did not work when the utility went down.