Optimizing NEM2.0 for PG&E EV2-A rate

[YRide]

PG&E takes the most advantageous (for them) interpretation: they remove the excess export credits from highest rate period first. I ran into this once when PG&E double-billed me by accident.

Ok, really good to know this. And yeah, kinda saw that coming. So I should really try to limit any export during off peak…if it looks like I’m going to clip the monthly maximums.

Though really no way to (easily) know this before the month closes i assume.

 

[aesculus]

Though really no way to (easily) know this before the month closes i assume.

You would have to download the data from the Tesla app for grid export and then fit it into your billing cycle dates. Not hard, but it will take a few minutes.

 

[Lloyd]

PACIFIC GRAFT AND EXTORTION!

 

[swedge]

PG&E takes the most advantageous (for them) interpretation: they remove the excess export credits from highest rate period first. I ran into this once when PG&E double-billed me by accident.

Ouch!

Not only did they take away those credits, but they also must have cost you hours of forensic accounting to decrypt the "detailed" bill to see what they actually did. ;-)

 

[Redhill_qik]

Very helpful, thank you.

I am on CCA (SVCE). Does this substantially change the math on things? Should I change to PGE to simplify things? I’m going to be offsetting most of my consumption, so I can sleep well at night (for the good of humanity). My understanding is the CCA v PGE production isn’t night a day anyways, wrt clean production.

I’ll be watching this stuff closely once I get PTO and can flip the two toggles “on”.

Looks like I should be ok to have headroom to add that split AC in ADU & electric dryer. Maybe next is swap gas heating for heat pump, that’s really the major one I suppose.

Btw, wrt VPP, that’s just a payment from Tesla direct, right? That doesn’t factor into all the NEM calculations does it?

Stay with CCA especially since it is SVCE. The accounting and billing statements have extra complexities and you will different annual true-up months between PG&E and SVCE, but if you are a net exporter then they pay twice the PG&E rates and the standard rates will be a little bit less, just don't select the 100% renewable GreenPrime add-on.

Before you start going down the path of trying to price arbitrage understand where you are without doing this. It can make sense to do this if your solar is undersized versus your usage, but if you expect to be close to $0 at true-up then grid charging and exporting can be a negative due to the 10% efficiency loss in the Powerwall.

 

[YRide]

Ok, I’m only a few days into having TBC + EE + PTO + GC all enabled. I think the system is still “training/learning”…but not sure it will get to what I need.

Right now (peak summer) I’m able to generate about 63kWh of PV a day. Without any home-load this would fill up my 51.3kWh ESS (2x PW+ and 2x PW2…54kWh @ 5% reserve). However, have normal home loads (inc AC) and these run during the day.

On PGE EV2A which has a 4-9p peak @ 56c/kWh, 3-4 & 9-12a shoulder at 45c/kWh and off peak at 26c/kWh. Sell = buy prices in both real life as well as in my Tesla app.

It seems that with the advanced settings above, the ESS starts GC @ 12a and goes full rate until it full with the ESS at 100% SoC, 54kWh.

Then my EV’s (two) start charging early AM (usually just after ESS are full) for a 745a departure target. All of this power comes from the grid, which makes sense, cheap off peak rate.

Then the sun starts coming up, and all of this power goes to the grid because the ESS are already at 100% SoC. This would be ideal IF pge did not cap generation to the grid based on PVWatts / Black&White bill. But since they do cap generation per @swedge on a monthly basis based on PVWatts, this is likely going to cause me to go over and per @cwied disallow my valuable peak generation NEM2 credits.

Then at 4p sharp (peak) the ESS dump at max rate down to 5% reserve, taking about 2.5hrs. This is slightly as expected, but I’m surprised it goes down to reserve without anticipating I have any any home load from 630-9p at peak and also 9p-12a at shoulder. Then, still in peak, abd then continuing to shoulder…I’m at reserve (so no more ESS) and the AC and other home loads are used…I’m drawing from the grid at peak rates, which I do not want to do.

I would think it would be financially beneficial to avoid those kWh’s either with time shifted PV generation into the ESS, or even off-peak GC to the ESS from earlier in the day.

So the issues I see are two-fold…

1) because of the PVWatts / monthly generation cap I clearly don’t want to GC the ESS to 100% SoC each night before the sun comes up. It will cause me to excessively generate to grid during the day from PV at off-peak rates. However, because I do have high consumption, both off peak (two EV’s) and off-peak/shoulder/peak/shoulder with things like AC…it would make sense for me to have some grid charging before the day starts so that ESS hits 100% SoC just before 4p (and then dumps itself to grid starting at 4p). Right now if I do GC off, it seems I can’t hit 100% SoC and this am leaving high valie NEM2 peak credits on the table. It doesn’t look like there is a way to make this happen is there? Seems like I need to leave GC off?

2) can I do anything (adjust buy/sell rates so they are not the same?) to drive ESS to not discharge all the way to reserve starting at peak (4p) so I have some ESS SoC above reserve remaining until midnight so I don’t have to pull any expensive peak or shoulder rate from the grid? Or is this just “learning/training” that will happen over the next few days?

Daily graphs attached.

Appreciate any thoughts to how to adjust settings to optimize, if at all possible.

 

[swedge]

Ok, I’m only a few days into having TBC + EE + PTO + GC all enabled. I think the system is still “training/learning”…but not sure it will get to what I need.

Right now (peak summer) I’m able to generate about 63kWh of PV a day. Without any home-load this would fill up my 51.3kWh ESS (2x PW+ and 2x PW2…54kWh @ 5% reserve). However, have normal home loads (inc AC) and these run during the day.

On PGE EV2A which has a 4-9p peak @ 56c/kWh, 3-4 & 9-12a shoulder at 45c/kWh and off peak at 26c/kWh. Sell = buy prices in both real life as well as in my Tesla app.

It seems that with the advanced settings above, the ESS starts GC @ 12a and goes full rate until it full with the ESS at 100% SoC, 54kWh.

Then my EV’s (two) start charging early AM (usually just after ESS are full) for a 745a departure target. All of this power comes from the grid, which makes sense, cheap off peak rate.

Then the sun starts coming up, and all of this power goes to the grid because the ESS are already at 100% SoC. This would be ideal IF pge did not cap generation to the grid based on PVWatts / Black&White bill. But since they do cap generation per @swedge on a monthly basis based on PVWatts, this is likely going to cause me to go over and per @cwied disallow my valuable peak generation NEM2 credits.

Then at 4p sharp (peak) the ESS dump at max rate down to 5% reserve, taking about 2.5hrs. This is slightly as expected, but I’m surprised it goes down to reserve without anticipating I have any any home load from 630-9p at peak and also 9p-12a at shoulder. Then, still in peak, abd then continuing to shoulder…I’m at reserve (so no more ESS) and the AC and other home loads are used…I’m drawing from the grid at peak rates, which I do not want to do.

I would think it would be financially beneficial to avoid those kWh’s either with time shifted PV generation into the ESS, or even off-peak GC to the ESS from earlier in the day.

So the issues I see are two-fold…

1) because of the PVWatts / monthly generation cap I clearly don’t want to GC the ESS to 100% SoC each night before the sun comes up. It will cause me to excessively generate to grid during the day from PV at off-peak rates. However, because I do have high consumption, both off peak (two EV’s) and off-peak/shoulder/peak/shoulder with things like AC…it would make sense for me to have some grid charging before the day starts so that ESS hits 100% SoC just before 4p (and then dumps itself to grid starting at 4p). Right now if I do GC off, it seems I can’t hit 100% SoC and this am leaving high valie NEM2 peak credits on the table. It doesn’t look like there is a way to make this happen is there? Seems like I need to leave GC off?

2) can I do anything (adjust buy/sell rates so they are not the same?) to drive ESS to not discharge all the way to reserve starting at peak (4p) so I have some ESS SoC above reserve remaining until midnight so I don’t have to pull any expensive peak or shoulder rate from the grid? Or is this just “learning/training” that will happen over the next few days?

Daily graphs attached.

Appreciate any thoughts to how to adjust settings to optimize, if at all possible.

Right! Try setting sell price 3 cents lower than buy. This is roughly what the Non-Bypassable Charges (NBCs) in NEM2 cost. These are billed on imports but not credited on exports. Including the offset in the app pricing will encourage PW to charge from solar rather than the grid. It will also make it avoid dumping too much to the grid too soon.

During winter rates mine guessed what my needs and tried to get to my 20% reserve at 12 am. Now that summer rates apply, it prefers to dump it down to 20% before 9 pm, preferring to export at the higher rate even though it means importing at the lower partial peak rate. It seems to get the math right, and even looks ahead based on weather forecasts of foggy days here in Oakland.

If the buy/sell are equal, it thinks there is no difference between solar or grid charging. Some folks have suggested that the round trip efficiencies of only 90% for charge/discharge cycles should impact the math too, but I have not tried to adjust for this yet.

Good work on paying attention to all these issues and PW behaviors!

 

[swedge]

But since they do cap generation per @swedge on a monthly basis based on PVWatts, this is likely going to cause me to go over and per @cwied disallow my valuable peak generation NEM2 credits.

With my settings, my export is usually around half of PG&E's estimated max. Every system and user is different, but I expect that if you set the buy/sell differential, your PW will do much more self-generation, and give you margin to keep your export below the estimated solar generation limit.

You can check this on each month's B&W detailed bill. The snail mail copies come weeks late, but .pdf's ar available online when your bill is created.

Log in to PGE.com, Hi,your name/My Accounts and Services/electric service id/Select from list to Download Detail of Bill.

 

[YRide]

With my settings, my export is usually around half of PG&E's estimated max. Every system and user is different, but I expect that if you set the buy/sell differential, your PW will do much more self-generation, and give you margin to keep your export below the estimated solar generation limit.

You can check this on each month's B&W detailed bill. The snail mail copies come weeks late, but .pdf's ar available online when your bill is created.

Log in to PGE.com, Hi,your name/My Accounts and Services/electric service id/Select from list to Download Detail of Bill.

Thanks. I’ll take a look, but it will be a month I suppose before I get the first one. I guess I can use PVWatts with my system design parameters to get a rough sense. But it would seem that barring lots of clouds, I won’t have that much headroom to TOU arbitrage + throw solar to the grid.

 

[YRide]

Right! Try setting sell price 3 cents lower than buy. This is roughly what the Non-Bypassable Charges (NBCs) in NEM2 cost. These are billed on imports but not credited on exports. Including the offset in the app pricing will encourage PW to charge from solar rather than the grid. It will also make it avoid dumping too much to the grid too soon.

During winter rates mine guessed what my needs and tried to get to my 20% reserve at 12 am. Now that summer rates apply, it prefers to dump it down to 20% before 9 pm, preferring to export at the higher rate even though it means importing at the lower partial peak rate. It seems to get the math right, and even looks ahead based on weather forecasts of foggy days here in Oakland.

If the buy/sell are equal, it thinks there is no difference between solar or grid charging. Some folks have suggested that the round trip efficiencies of only 90% for charge/discharge cycles should impact the math too, but I have not tried to adjust for this yet.

Good work on paying attention to all these issues and PW behaviors!

I don’t understand why in your case for summer or would want to deplete by 9p? Are you on EV2A or some other TOU plan?

With EV2A it would make no sense to consume grid at either peak, when you could either use time delayed Solar (or time delayed off peak grid charge). So the fact mine is depleting ESS well before peak ends doesn’t make sense.

I guess it was finishing right at 9p then round trip loss on arbitrage from shoulder to peak does make sense. That’s 10% loss on 56-45c which is 11c-1.1c=9.9c/kWh.

Still think it would be better to hold a little extra charge in the ESS, then recover that via grid charge at 12a next day.

I just updated sell pricing per your guidance. Let’s see how that goes.

 

[swedge]

Thanks. I’ll take a look, but it will be a month I suppose before I get the first one. I guess I can use PVWatts with my system design parameters to get a rough sense. But it would seem that barring lots of clouds, I won’t have that much headroom to TOU arbitrage + throw solar to the grid.

Your tweak of buy/sell prices should fix this. Probably already has, since it is now tomorrow. ;-)

My PW is not doing any grid charging these days, 'cause even with the foggy mornings here it knows it'll get fully charged before 3. Usually it is 100% before 1PM.

Some days I've been charging our Model Y after PW gets to 100% till 3pm to reduce the NBC hit on exports. This saves less than a buck a day, but then 365 is a significant multiplier... Supposedly there is a "charge on solar" option coming soon for Tesla cars, but for now it is a manual thing.

My first NEM2 true-up just came in with only a ~$65 "adjustment", so I seem to be doing something right...

 

[YRide]

Your tweak of buy/sell prices should fix this. Probably already has, since it is now tomorrow. ;-)

My PW is not doing any grid charging these days, 'cause even with the foggy mornings here it knows it'll get fully charged before 3. Usually it is 100% before 1PM.

Some days I've been charging our Model Y after PW gets to 100% till 3pm to reduce the NBC hit on exports. This saves less than a buck a day, but then 365 is a significant multiplier... Supposedly there is a "charge on solar" option coming soon for Tesla cars, but for now it is a manual thing.

My first NEM2 true-up just came in with only a ~$65 "adjustment", so I seem to be doing something right...

Ok. I’ll keep watching. For my system on the summer solstice with no clouds I’m only generating 63kW with PV and so will not charge 51kWh ESS until later in day.

So it sounds like you have a bigger PV/ESS ration then I do? I’m 0.2x (10.8kW/54kWh)

 

[hiltxan]

I'm also trying to learn how to adjust the buy/sell rates as well as time for off peak, partial peak and peak to maximize savings/earnings while also balancing charging EV vehicles and keeping powerwalls to a certain percentage.

Sometimes the buy/sell rates just seem to confuse the user and the gateway trying to be "too smart." I wish they got rid of the buy/sell rates and just had priorities instead for where to export solar and where to import for powering the home. For instance, during off peak times, there should just a way to set where you want your power to come from - grid, powerwall (to a certain %) or both. Then during the day, I also run into problems where solar will charge the powerwall first while also drawing power from the grid (thus incurring NBCs). What's even more annoying is after I adjust all the rates, it will work as my intentions for half a day and then revert back to what it was doing before.

 

[BGbreeder]

Right! Try setting sell price 3 cents lower than buy. This is roughly what the Non-Bypassable Charges (NBCs) in NEM2 cost. These are billed on imports but not credited on exports. Including the offset in the app pricing will encourage PW to charge from solar rather than the grid. It will also make it avoid dumping too much to the grid too soon.

During winter rates mine guessed what my needs and tried to get to my 20% reserve at 12 am. Now that summer rates apply, it prefers to dump it down to 20% before 9 pm, preferring to export at the higher rate even though it means importing at the lower partial peak rate. It seems to get the math right, and even looks ahead based on weather forecasts of foggy days here in Oakland.

If the buy/sell are equal, it thinks there is no difference between solar or grid charging. Some folks have suggested that the round trip efficiencies of only 90% for charge/discharge cycles should impact the math too, but I have not tried to adjust for this yet.

Good work on paying attention to all these issues and PW behaviors!

I think that a sell price of more than the NBC is indicated to allow for the Powerwall charging losses (90% round trip efficientcy), so the sell price in the app should be less than 90% of the EV2A rate minus NBC to accurately account for the true costs/price of the energy.

All the best,

BG

 

[aesculus]

I think that a sell price of more than the NBC is indicated to allow for the Powerwall charging losses (90% round trip efficientcy), so the sell price in the app should be less than 90% of the EV2A rate minus NBC to accurately account for the true costs/price of the energy.

All the best,

BG

That is certainly true if you are exporting from the PW, but not if you are just exporting solar directly.

It would be super nice if Tesla had a setting for PW export but we are certainly in the early stages of that scenario so TBE.

 

[h2ofun]

Your tweak of buy/sell prices should fix this. Probably already has, since it is now tomorrow. ;-)

My PW is not doing any grid charging these days, 'cause even with the foggy mornings here it knows it'll get fully charged before 3. Usually it is 100% before 1PM.

Some days I've been charging our Model Y after PW gets to 100% till 3pm to reduce the NBC hit on exports. This saves less than a buck a day, but then 365 is a significant multiplier... Supposedly there is a "charge on solar" option coming soon for Tesla cars, but for now it is a manual thing.

My first NEM2 true-up just came in with only a ~$65 "adjustment", so I seem to be doing something right...

I do not touch my system at all, and get a huge adjustment. I must be doing something right?

 

[swedge]

Ok. I’ll keep watching. For my system on the summer solstice with no clouds I’m only generating 63kW with PV and so will not charge 51kWh ESS until later in day.

So it sounds like you have a bigger PV/ESS ration then I do? I’m 0.2x (10.8kW/54kWh)

My system is not as large as yours. I think my signature text says it. 6kW solar, 1 PowerWall. NEM2, EV2-A

Now, near the solstice, on a cloud free day the solar produces around 30 kWh. My house has (mostly) no air conditioning, gas heat. Daily house consumption is around 15kWh. This last year I've largely stayed set on Export Everything and Grid Charging, with a 20% backup reserve.

The car averages a similar amount, but not every day, charge starting after the midnight rate drop.

I sized the solar and setting to try and keep the true-up adjustment near zero. As I wrote earlier, I just got my first NEM2 true-up with an adjustment of $67.43.

Again, every situation is different. Climate, solar sizing, HVAC type, storage capacity, car use each have large impacts. If we ever get around to replacing the gas furnace, water heater and clothes dryer with heat pumps, we're gonna need a bigger boat... uh.... make that solar and battery system.

 

[swedge]

I do not touch my system at all, and get a huge adjustment. I must be doing something right?

Indeed you are. If we had the budget, we'd replace gas with heat pumps, make the solar as large as the roof, and add more storage.

I am quite please to be able to charge our car without blowing up the PG&E bill. I think the new 7 additional solar panels which enabled this cost us $7k, less the tax credit. So it wasn't free.

 

[swedge]

I think that a sell price of more than the NBC is indicated to allow for the Powerwall charging losses (90% round trip efficientcy), so the sell price in the app should be less than 90% of the EV2A rate minus NBC to accurately account for the true costs/price of the energy.

All the best,

BG

That does sound like a better approximation. But I do not rely on the $$ numbers from the Tesla app, I just want the PW to behave "correctly".

On my EV2-A rates, the discounted partial peak export would still come in higher than off-peak import, and peak export even more. So I would not expect your adjustment to effect the PW export or grid charging behavior.

Or am I missing something here?

 

[BGbreeder]

That is certainly true if you are exporting from the PW, but not if you are just exporting solar directly.

It would be super nice if Tesla had a setting for PW export but we are certainly in the early stages of that scenario so TBE.

FWIW, I'm not exporting from my powerwalls.(Not allowed, except for peak demand events.) I did not get reasonable export behavior from my solar/Powerwall system until I modified the pricing, but I am entirely willing to believe it was coincidental.

All the best,

BG