Understanding Why Tesla Energy Settled on 7kWh Powerwall

AUG 27 2015 BY MARK HOVIS 38

Solar Calendar

Solar Calendar – a year in a glance (graph provided by Enphase Enlighten interface)

There has been a lot of speculation on just how solar users will select the right size energy storage for their application, particularly in reference to the 7kWh size of the Tesla Energy Powerwall. For customers who have access to net metering, (the practice of selling excess solar energy at the same price as one purchases energy), there currently is little incentive for energy storage. So for the moment, the U.S. market for residential solar backup is limited.

For those that are exposed to Time of Use (TOU) rates, or wholesale pricing like that offered by energy co-operatives, or those who want to understand how residential solar could play a critical role in distribution of balancing peak load demands, then this attempt of explaining through an applied data approach may help in understanding how the Tesla Energy 7kWh Powerwall, and similar sized storage systems, could be integrated.

The above graph is a shaded representation of a 6kW DC rated manual tilt solar array located near Charlotte North Carolina receiving on average 5kWh/m2/day. The lighter data represents premium solar activity, while the darker indicates minimum solar production. Below are two sets of data. The first being the numeric kWh daily generation totalling 8860 kWh. On average, this array sells 6000 kWh per year from daytime excess, and purchases 6000 kWh for evening an inclement usage.

The second table is revealing. On average,  16 kWh are used daily directly from the array except during southern summer months where 25 kWh are directly used. The second table subtracts an average 16 kWh from the generated amount and 25 kWh for summer months.

Solar Data

Annual Solar Data mirrors top graph

Here is the takeaway. The displayed data shows how many days could be used for solar storage and how much. The orange shaded data reveals excess under 9kWh. The green indicates the potential for two Powerwalls or comparable energy storage. Non shaded data reveals days with no storage capacity.

This displays on one hand how difficult planning an off grid system would be, while on the other hand, how successful energy storage could be used in reducing the duck curve, or peak utility load by distributing power back to the grid for  4-5 hours. The latter, being the the real potential for residential solar storage.

Duck curve

Duck Curve

Without energy storage, solar only offsets peak load until 3:00 – 4:00 PM. The ideal solution would be to distribute energy back to the grid from storage for an additional four – five hours. System size and usage varies greatly. As solar adoption grows, it will reach utility capacity faster than say if future systems are balanced with energy storage. It is difficult for U.S. individuals to visualize the need while they are being rewarded with the net metering model.  The imbalance of peak loads, combined with the ability of an energy storage solution, will drive a future model that will likely be designed around this burgeoning industry.

Until utilities or government are willing to pay a premium for the stored energy, it will be difficult to achieve this elegant solution to peak load.

Night charging offers new energy market

Night charging offers new energy market

After 10:00 PM, the duck curve, or utility generation goes very flat. Night time EV charging will open a more balanced market for utilities.

In addition to managing peak load is adding grid stability. The key, Nahi said, is that data collected by modern inverters and potentially by other parts of PV systems can be used to help utility companies or transmission and distribution (T&D) operators to manage networks.” Enphase CEO, Paul Nahi, states  that data collected by modern inverters and potentially by other parts of PV systems can be used to help utility companies or transmission and distribution (T&D) operators to manage networks.

This leaves us with what to do with long periods of intermittent weather. This final piece very well may be filled with utility scale power packs, and/or other methods of gathering excess renewable energy such as the use of  hydrogen and similar methods which combined will bring the most balanced grid we have known to date.

The first phase of solar adoption has already begun with the rapid expansion of utility scale and residential solar. The next phase will include the addition of energy storage. Currently, utilities are making up the bulk of storage purchases and may very well have realized a future in energy storage.  If the utilities choose to continue the offering of net metering, they very well will control energy storage indefinitely. If net metering goes away, the market will respond with an answer to both peak load and individual economics, and Tesla Energy may have sized it just right

Categories: Tesla

Tags: , , , ,

Leave a Reply

38 Comments on "Understanding Why Tesla Energy Settled on 7kWh Powerwall"

newest oldest most voted

Good luck integrating that Enphase based system with the Tesla Powerwall. :-/ (I have the same issue. Enphase should build an AC coupled battery front-end for the Powerwall if they have any brains.)

Probably depends on the definition of “integration”

I think the Powerwall will work just fine for what it’s intended to do, but the traditional Enphase inverters will shut off if external grid power goes away.

At least in my agreement with National Grid and NYSERDA, I agreed *NOT* to use battery storage as long as I’m benefiting from Net Metering.

And company wide, its limited to 1% of the residence total energy sold to residence customers. Commercial and Industrial have their own agreeements.

This is curious to me – could you elaborate? What does National Grid lose if you use a battery with net metering? Are you on a time-of-use schedule? Or do they recognize that the timing of solar power is near ideal for offsetting peak usage, and don’t want to lose that benefit?

This actually applies to any investor-owned utility in NY State. Its not NG specific.

There’s no ‘harm’ done to NG in any event, since any program limited to 1% participation is by definition small potatoes.

Nope, I have ths plain old SC-1 service class, as I’ve mentioned I use so little electricity actually that time-of-use would be more costly for me. I pay $15.67 each month, but currently am working off a $155 credit since I was given $.0357 per kwh on my annual overage (I was 4344 kwh over last year, and thats with a whopping 1400 kwh increase in usage from 2013 to 2014. – 2013 I used roughly 3500 kwh per annum, and last year I used about 4900 kwh, but I made over 9300 kwh from the panels). Depending on the upcoming winter, I expect to use less this year, since preheating the Roadster battery all the time used at least a 1000 kwh over the season. The volt and elr require much more minimal preheating and is just a small cost in charging. But since the engines run in the cars when its cold, I’ll use much less electricity anyways, except that 36 degree engine turn on point will drive me crazy. I’ve alredy identified which 2 fuses to pull from the elr and i’ll do it while its under warranty to see if i can keep the engine from… Read more »

They have their own “AC Battery”.

I’m not saying it’s good or bad, but they have leveraged their small inverter technology and put it with a battery. Quite small modules though – 275W / 1.2kWh – that can be daisy chained together to make the size you want.

I don’t know anything about availability or price.

Well . . . they’ve announced their own AC battery anyway. But the details are vague. It is not available. And it is ridiculously small.

If Enphase wants to do it right, they need to build an AC front end to the Powerwall that will disconnect from the grid when there is a grid failure and generate an AC signal that will keep their microinverters operating.

But so far, it seems they don’t get it such that their stock has crashed.

Monolithic inverters of multiple kilowatts is really not their thing, so why would you want THEM to do it? I can see how the one inverter per panel thing has benefits. I don’t see how having such small inverters and battery packs is helpful.

BTW, that duck graph is a few years old now. I’d like to see an updated version of it. I wonder if someone has one.

Is that a Model X I see in the bottom picture, surrounded by a bunch of people? And no one’s taking spy pics of it?


The duck curve is a quack.

It represents a best case scenario for the utilities fight against distributed energy and rates favorable to solar customers in California, a bright March day with no need for an air conditioner.

The graph on a summer’s day with heavy load and solar generating into the 6pm hour is vastly different. I guess we could call that a pregnant duck.

That’s the load the utilities need to service.

Yeah, I want to see an updated Duck graph. And I think they can solve the problem with a combination of super-peak prices at that neck time, demand-response, getting PV installers to face more solar PV to the west, thermal storage like AC systems that can run on ice they created at night, and bit of other storage systems.

Nice write up Mark, thanks!

With the demand curves I’ve seen for my area, solar does as you say, offsetting much of the peak until around 4pm. However, I would think the only major “gap” is from 4-5pm, since after 5pm, many businesses will close and their power usage will decrease. Granted, people will go home and start using power there, but I’m guessing that the main demand is with the commercial locations.

I’ve always liked how solar is relatively aligned with power demand. With a little help from battery storage on the shoulder, it seems like we can have a very stable grid.

V2G technologies still seem very promising in this regard as well, as EVs become more proliferated.

I’d like to see the case of where an owner willingly increases his battery wear and tear for the amount of money a utility would willingly pay.

Outside of California’s somewhat scam system , and much of the Northeast’s confiscatory rates, that is.

I know there’s been dozens of articles on this, but if its so compelling, the utilities would do it themselves, which they do.

Especially around wind farms where the output is choppy enough that flywheel generator systems, pumped storage systems, on battery systems ameliorate the huge wind farm outputs, with the pumped storage systems being the most cost effective, if available.

One of the many problems with wind farms is that their output is so variable in such a short period of time, and so unpredictable, that their power output can seriously destabilize an entire region of the electrical grid. Hence utilities have a strong motive to install power storage to even out the power coming from wind farms. But for such regulation, only a few minutes of storage is needed to prevent destabilizing the grid. This is a long way from the potential, which would be several hours of storage so the utility could perform the same sort of energy “time shifting” over a 24 hour period as this article discusses. However, Bill, I think you raise a valid point, which all too many discussions of consumer use of battery backups lack. The point is this: If it’s good for the consumer, then why isn’t it even better for the utility? Why don’t utilities install lots of battery backup capacity, to benefit from the very same cost/benefit which is now benefiting many home owners and commercial customers? Of course, it’s not at all simple. It depends partly on the rate structure for selling electricity, and that may well be out… Read more »

As I understand the situation, power utilities are heavily regulated, whether public or private. They simply don’t have the flexibility to save up capital over a couple years and then spend the money needed to provide modern battery buffering systems.

Yeah, the only pumped storage in my area is in Niagara falls, and since its the cheapest method, can be used for ballancing out loads. We don’t have any huge wind farms in my area anyway.

NYSEPA or whatever they call themseleves is putting enough silly restrictions on our one sole remaining 380 megawatt coal-fired (used to be around 1200 megawatts before its basic destruction over the years) – namely they can’t run it at over 60 megawatts, that NRG has finally ‘had it’ and is closing the plant entirely.

In NYS the ‘utility’ no longer generates the power, so such things are of no concern to them, but thermal fatique or turbine MELTING due to Joule-Thomson issues at reduced output, in addition to thermal fatigue issues is of GREAT concern to the generating companies, and it would not surprise me to see 100 MW sized flywheel storage or battery plants owned by the generators simply to minimize damage to their plants, and indirectly allow for more efficient operation of them.

V2G is already great and growing. If you have 20+ year battery life with liquid cooling like my FORD FOcus EV does along with those with a Volt the battery needs to be used for a big payback with V2G, V2H.
Read www/V2G-101.webs.com

The “Duck Curve” presented above is misleading, as this particular instance only represents a day in Apr, or Sep.

The shape of the solar energy and grid energy morphs over each month of the year. eg: during Jun-Aug, higher A/C use is offset by longer solar energy production. In winter Nov-Feb the grid is much flatter, as wind complements periods when solar production is low.

Another issue with duck curve, it assumes energy is only produced in one time zone and doesn’t flow east-west across time zones to areas where there is greater demand. eg; when AZ, CA have lower morning/mid-day demand … east-coast states can use the extra energy.

IMO: Don’t place too much analysis on a single duck curve … at a minimum be sure to consider the local energy pattern for each quarter of the year, or better each month of the year. The grid, and your local energy use does vary!

Well, they do have to build the grid for the worst case situation.

And the East-west bit doesn’t help much since there is nothing to the west of us but the Pacific ocean. (Of course, putting some wind turbines out there would be nice.)

The US is broken into three main grids – east (Oklahoma and everything north and east), west (New Mexico and everything north and west), and Texas (ERCOT).

You’d have to build a way to move power between the grids in a way that is also fault tolerant and can prevent cascading brownouts and blackouts.

Nice write-up Mr. Hovis, Thanks!

Where in the U.S. does net metering mean you get the same price as you would pay? At least in California PG&E only gives me 4c per kWh that it would charge me 30c for, the so called wholesale vs retail price difference – thankfully only on excess difference after 12 months aggregate, but still a huge discount for them when buying electricity from me compared to what they have charged me for years before I got a PV system.

PG&E gives you 4 cents/KWH for the EXCESS above what you use for the entire year.

That’s when that 4 cents comes into play. For the excess during the day that they later pay back with electricity at night, I think they use their standard rates.

In the PG&E area in N. California you get full retail for energy fed back to the grid. If you provide more than you consume then you get avoided costs for the excess power you feed back. That is about $.04 per kWh.

With the powerwall; what stops one from CHARGING at night cheap; and DISCHARGING to the grid during the day? Charge at 0.12 kwh and discharge at 0.49 kwh?

Went we went net metering, we specifically had to agree to go non-battery for grid-tied systems to avoid such gaming scenarios.

SCE in Southern California has true net metering calculated monthly and then balanced over a year.

Typo in the third paragraph:

“The above graph is a shaded representation of a 6kWh manual tilt solar array”

… should probably say:

“The above graph is a shaded representation of a 6kW manual tilt solar array”

(kW vs kWh)

Yes, it matters 🙂

You might add a sentence explaining that the 6 kW is the DC rating of the collective panels, and that the peak AC output of the system is about 4.8 kW (80% of DC rating).

Finally, I would credit Enphase for producing that neat checkerboard graphic at the top. I have that Enlighten interface for my own Enphase microinverter system and it’s pretty cool.