busbar ratings with an ESS

[wwhitney]

You are missing, as I've been trying to explain above, that grid export (or lack thereof) isn't the paramount factor.

But to be fair, an ESS that is "backup only" shouldn't ever discharge while the grid is up, so there are no panels that have more than one simultaneous source of supply. {PV being a separate issue.] Whether that should satisfy 705.12 without the ESS being listed as a PCS is an interesting question.

Cheers, Wayne

 

[rainwater01]

But to be fair, an ESS that is "backup only" shouldn't ever discharge while the grid is up, so there are no panels that have more than one simultaneous source of supply. {PV being a separate issue.] Whether that should satisfy 705.12 without the ESS being listed as a PCS is an interesting question.

Cheers, Wayne

That’s certainly fair. The Generac has a “self consume” mode where it’ll prioritize PV and batteries on the household loads before it’ll pull anything from the grid. The customer wanted to use that. I told her I would turn that off because the grid costs 12 cents a kwh and the batteries are like 22 cents a kwh to get it through the warranty period. So it’s possible but makes no sense to do so.


Sent from my iPhone using Tapatalk

 

[jaggedben]

But to be fair, an ESS that is "backup only" shouldn't ever discharge while the grid is up, so there are no panels that have more than one simultaneous source of supply. {PV being a separate issue.] Whether that should satisfy 705.12 without the ESS being listed as a PCS is an interesting question.

Cheers, Wayne

Most systems have user settings to change modes, as rainwater describes. Per the code and (I believe) UL, these settings can't be relied on as PCS setpoints; the setpoints must only be available to the installer and manufacturer. So no, that would not satisfy the code.

 

[wwhitney]

Per the code and (I believe) UL, these settings can't be relied on as PCS setpoints; the setpoints must only be available to the installer and manufacturer.

Agreed that 705.13(E) would require a "backup-only" setting in the installer menu, rather than in the user menu.

Cheers, Wayne

 

[jaggedben]

That’s certainly fair. The Generac has a “self consume” mode where it’ll prioritize PV and batteries on the household loads before it’ll pull anything from the grid. The customer wanted to use that. I told her I would turn that off because the grid costs 12 cents a kwh and the batteries are like 22 cents a kwh to get it through the warranty period. So it’s possible but makes no sense to do so.


Sent from my iPhone using Tapatalk

That's flawed reasoning. The battery is a sunk cost that she's already paid for. So if her solar cost per kwh is enough less than 12cents to cover the battery round-trip efficiency, she can still save money. Of course at those rates it won't be much, and it won't cover the cost of the battery. But she'll end up paying less out of pocket than otherwise. On PG&E'S EV rate where peak is 49cents and off peak is 19cents, the battery can actually compete over the long term.

 

[rainwater01]

That's flawed reasoning. The battery is a sunk cost that she's already paid for. So if her solar cost per kwh is enough less than 12cents to cover the battery round-trip efficiency, she can still save money. Of course at those rates it won't be much, and it won't cover the cost of the battery. But she'll end up paying less out of pocket than otherwise. On PG&E'S EV rate where peak is 49cents and off peak is 19cents, the battery can actually compete over the long term.

Maybe you're right. What part of this am I not understanding correctly? Maybe you're thinking of the battery as being paid for so it doesn't count in the equation and I'm thinking of the battery as prepaid electricity that gets lower every time you use it.

I figure that if the battery costs $1,663.29 and has a 10 year 7.56MWH throughput then the energy you purchased costed you $.22 per kwh.

So if you use 10kwh of energy from the battery at night it'll cost you 10kwh x $.22= $2.20
Then the next day instead of selling 10kwh x $.06 = $.60 you're charging your battery instead.
Therefore your cost was $2.80 for that 10kwh.

On the other hand if you bought that power at night from the grid for 10kwh x .12 = $1.20
Then the next day instead of charging your battery you sold that amount of power back for 10kwh x $.06 = $.60
So you're total cost would be $1.20-$.60 = $.60

So if I use the battery I'm paying $2.20 and if I use the grid I'm paying $.60

If I have (6) 7.56MWH batteries and use 12.43kwh per day then I can expect the batteries throughput to last through the 10 year warranty period.

If you ignore the price of the battery in the calculation then it doesn't cost you anything but you should have a sinking fund for when the batteries need replaced.

I figure as long as she produces twice the amount of power as she buys then the only fee she'd have is the "meter fee" but if the battery doesn't get used and dies of old age then like anything else in life it's worth nothing because you bought it and never used it. There must be a happy median in there somewhere.

Let me know if I'm thinking crazy because I respect your opinion. I'm thinking of the batteries as more of a backup power source rather than a money saving device.

Thoughts?

 

[jaggedben]

Maybe you're right. What part of this am I not understanding correctly? Maybe you're thinking of the battery as being paid for so it doesn't count in the equation and I'm thinking of the battery as prepaid electricity that gets lower every time you use it.

I think you need to more thoroughly develop your thinking.

I figure that if the battery costs $1,663.29 and has a 10 year 7.56MWH throughput then the energy you purchased costed you $.22 per kwh.

That's a useful figure to compare if using the battery at full capacity to store solar can compete with utility rates . You'd add $.22 to a similar number for the solar production and see if it competes with utility import cost. (Most places it can't, yet.) But that's not why she bought the battery, right? The way you should think of it is that she paid $1,663.29 for a backup power source, but if can she use that same system to save money on the utility bill, then over time she ends up paying less than $1663.29 for the backup power source.

Also (and this is a fine point but helpful to understand) but she didn't purchase any energy with the money for the battery. The energy is obtained a different way, mainly from PV.

So if you use 10kwh of energy from the battery at night it'll cost you 10kwh x $.22= $2.20

Nope, you already paid for it. Using it doesn't increase or decrease the amount you paid for it.

Then the next day instead of selling 10kwh x $.06 = $.60 you're charging your battery instead.
Therefore your cost was $2.80 for that 10kwh.

On the other hand if you bought that power at night from the grid for 10kwh x .12 = $1.20
Then the next day instead of charging your battery you sold that amount of power back for 10kwh x $.06 = $.60
So you're total cost would be $1.20-$.60 = $.60

So if I use the battery I'm paying $2.20 and if I use the grid I'm paying $.60

Nope, this is wrong. Let's consider...

So grid imports cost 12cents, and grid exports get credited at 6 cents. We need to know what the PV costs (cost to install divided by lifetime generation), but let's say 8 cents for the sake of example, which is not unrealistic.

So now, if she stores 8cents per kWh worth of solar in the battery, then uses it later to avoid 12cents of grid imports, she's saved 4 cents on avoided grid imports. Okay let's say 3cents because of battery efficiency. Not only that, but she's no longer effectively losing 2 cents on solar exports. So she's actually saving 5cents on each kWh cycled through the battery, that comes off her utility bill. (Well, maybe not that much if there's a 'mimimum bill', but perhaps that much.)

So perhaps if she maximizes throughput of the battery she can save 5/22 of the battery cost over 10 years.
That has to be weighed against the intangible value of having the battery fully charged in an unexpected outage.

I figure as long as she produces twice the amount of power as she buys then the only fee she'd have is the "meter fee" but if the battery doesn't get used and dies of old age then like anything else in life it's worth nothing because you bought it and never used it. There must be a happy median in there somewhere

Let me know if I'm thinking crazy because I respect your opinion. I'm thinking of the batteries as more of a backup power source rather than a money saving device.

Yeah, I think the happy medium is basically that the batteries are a backup source that might be able to recoup some of their cost to install. (Unlike fuel generators, which won't recoup the cost. Also a fuel generator could also get paid for and never used before it goes bad.)

The economics in your area may differ, but in my area I can tell people "The solar can pay for itself, and for the battery. The battery can also pay for a bit of itself. So over 25 years you save money and you get a backup source for 10-15 years."

 

[wwhitney]

So grid imports cost 12cents, and grid exports get credited at 6 cents. We need to know what the PV costs (cost to install divided by lifetime generation), but let's say 8 cents for the sake of example, which is not unrealistic.

I don't think you need to know the PV cost at the point you start operating, it's a sunk cost. You would need to know the PV cost to decide how much PV is worth buying.

But once the PV and ESS are operating, there's 3 options for each kWh it produces:

a) reduce grid consumption by 1 kWh immediately, that's worth 12 cents.
b) If PV production exceeds house demand, then export the excess kWh, that's worth 6 cents.
c) put an excess kWh into the battery, and use it to offset a future (e.g. night-time) 0.9 kWh (due to round trip losses). That's worth 0.9 * 12 cents = 10.8 cents.

So (c) is better than (b), for the numbers you made up.

Cheers, Wayne

 

[jaggedben]

Agreed, the PV cost is only needed in my example to show how it might affect the long term net cost of the system.

 

[rainwater01]

I think you need to more thoroughly develop your thinking.



That's a useful figure to compare if using the battery at full capacity to store solar can compete with utility rates . You'd add $.22 to a similar number for the solar production and see if it competes with utility import cost. (Most places it can't, yet.) But that's not why she bought the battery, right? The way you should think of it is that she paid $1,663.29 for a backup power source, but if can she use that same system to save money on the utility bill, then over time she ends up paying less than $1663.29 for the backup power source.

Also (and this is a fine point but helpful to understand) but she didn't purchase any energy with the money for the battery. The energy is obtained a different way, mainly from PV.


Nope, you already paid for it. Using it doesn't increase or decrease the amount you paid for it.

Nope, this is wrong. Let's consider...

So grid imports cost 12cents, and grid exports get credited at 6 cents. We need to know what the PV costs (cost to install divided by lifetime generation), but let's say 8 cents for the sake of example, which is not unrealistic.

So now, if she stores 8cents per kWh worth of solar in the battery, then uses it later to avoid 12cents of grid imports, she's saved 4 cents on avoided grid imports. Okay let's say 3cents because of battery efficiency. Not only that, but she's no longer effectively losing 2 cents on solar exports. So she's actually saving 5cents on each kWh cycled through the battery, that comes off her utility bill. (Well, maybe not that much if there's a 'mimimum bill', but perhaps that much.)

So perhaps if she maximizes throughput of the battery she can save 5/22 of the battery cost over 10 years.
That has to be weighed against the intangible value of having the battery fully charged in an unexpected outage.

Yeah, I think the happy medium is basically that the batteries are a backup source that might be able to recoup some of their cost to install. (Unlike fuel generators, which won't recoup the cost. Also a fuel generator could also get paid for and never used before it goes bad.)

The economics in your area may differ, but in my area I can tell people "The solar can pay for itself, and for the battery. The battery can also pay for a bit of itself. So over 25 years you save money and you get a backup source for 10-15 years."

Thanks for the comments. What I meant by the prepaid electricity is that I’m buying a storage device that’s life is shortened every time I use it. Once it’s throughput is used up I have to buy another one to store the electricity I produce all over again. Of course if the batteries last considerably longer than the warranty period then the ROI is higher and therefore makes the electricity I’m pulling from them less expensive. The higher your energy costs, like in your area, the more viable it is to use the batteries to save or make money. For me I would want the battery to last as long as possible and my investment in batteries would be peace of mind knowing that I have electricity when the grid is down.

I consider you and Wayne as geniuses so I’m going to read your posts over again until I wrap my brain around it better If you both agree I’m sure you’re right.

How long do you guys think lithium batteries will last beyond the warranty? Will time or throughput kill them?


Sent from my iPhone using Tapatalk

 

[wwhitney]

What I meant by the prepaid electricity is that I’m buying a storage device that’s life is shortened every time I use it. Once it’s throughput is used up I have to buy another one to store the electricity I produce all over again.

That is a valid point, capacity of the ESS generally depends on both calendar aging and cycle aging. And at some point the capacity will hit a low enough value (70%? 50%? of original capacity) that you will call it end of life and consider replacing it.

So the question is how to model capacity degredation. There are two simple models: one is that calendar aging is immaterial, and capacity degradation depends solely on kWh cycled through the battery. For that model, if you know the kWh through-cycled capacity to end of life, you can divide the ESS cost by that to come up with an additional cost per kWh for my option (c). And then the question is whether (c) is still a better deal than (b). That model is basically what you initially proposed.

The opposite extreme model is that calendar aging is all that matters. Then the marginal battery cost of cycling a kWh through it is 0. That is the model jaggedben and I assumed.

Reality is somewhere in between, so a more sophisticated analysis would be required to be more accurate. Unless one can say with confidence reality is pretty close to the two endpoint models, and then just use that model. I'm not sure.

Cheers, Wayne

 

[jaggedben]

Thanks for the comments. What I meant by the prepaid electricity is that I’m buying a storage device that’s life is shortened every time I use it.

Yeah, I think this is pretty far of the mark. Not that there's nothing to it. But I mean, it is definitely not the case that your battery will last 30 years instead of 10 years if you use it 1/3rd as much. Like Wayne I don't have the knowledge to model it in detail. Also it will depend on the exact battery chemistry and product, and is probably trade secret to a large degree. To the best of my understanding, for lithium batteries anyway, regular not-too-deep cycling is the healthiest for the battery. Routinely discharging to zero is bad, but leaving fully charged forever isn't good either. There's a reason that manufacturer's warranties have e.g. a 10 year expiration even if you don't reach the warrantied throughput. So finding the most economical 'backup reserve point' for the battery probably requires a sophisticated model of the battery, as well as the economics of utility interaction.

 

[GoldDigger]

If the primary motivation for the system is as backup power, you need to know the reduction in battery life with each discharge cycle to decide whether to also use it for time shifting.

Sent from my Pixel 4a using Tapatalk

 

[rainwater01]

I just ordered 9Kwh of batteries for my own system. I figure if Generac is confident enough that they’ll warranty them for 10 years or 22.68Mwh I’ll aim at using 10% of that throughput per year.

I’ll let you know in 10 years how it works out for me


Sent from my iPhone using Tapatalk

 

[pv_n00b]

Hmmm....

First, so far none of the PCS systems I've worked with allow you to do that. They all monitor the system output with CTs at a certain point and must limit the output at that point. They cannot monitor the input to a loadside busbar while also monitoring the grid iport at the same time and make the latter part of PCS operations. I don't see why it isn't possible to do what you describe, but so far it does not seem 'typical' to measure grid imports as part of PCS operation.

Second, thinking like an overly conservative member of a CMP, you could still have a scenario where the system is configured to measure the house load and output battery to cover it, and the house load goes above the busbar rating, and the battery, PV, and utility are all feeding more into the busbar (from opposite ends, to be sure) than it's rated for. This has be prevented either with sufficient busbar rating or PCS. It's the same as the situation we've always had with just solar, just taken to another level of extreme with ESS. But this point is all this is still the case without having to do with the ESS exporting to the grid.

A PCS can be set up to monitor whatever someone wants to set it up to monitor. The 2020 NEC 705.13 spells out how the PCS needs to monitor and control the loading of conductors and busbar.

705.13(A) Monitoring . The PCS controller shall monitor all currents within the PCS. Any busbar or conductor on the load side of the service disconnecting means that is not monitored by the PCS shall comply with 705.12. Where the PCS is connected in accordance with 705.11, the PCS shall monitor the service conductors and prevent overload of these conductors.

It can only do that if it monitors all the supplies to a conductor or busbar including utility, PV, BESS, etc., not just how much one source is supplying. There are no standalone PCS systems on the market that does what 705.13 requires that I know of so I can see how you have not worked with any. Most larger C&I BESS systems monitor multiple points in the system and control the output of the BESS to prevent overload by the BESS when added to other sources.