Is Enphase good/worth cost for microinverters for small system? Why don't many people do smaller systems?

[1201]

Funny story, I tried using my power company website however they have some bad javascript (they just updated to angular/SPA design), so nearly all of their usage reports are broken/don't work at all, so I have resorted to the old fashioned way of kW monitoring until they fix it.

I was researching some hybrid inverters, the MP LV6548 looks nice but I'm not sure I could zero export with it, I'll need to do some more research. At this point, doing a small system with offset and hybrid inverter, and maybe a small battery might have a good payoff for me. Like you said, trying to schedule the battery for usage in demand hours after the sun is down would be ideal.

Emporia vue-.

Lv6548 is zero export since it's off grid. You'll need two.

Youll need to land the loads you intend to power into a new panel, or power the main panel with an interlock switch and land any loads you don't want to power to a new sub panel.

You can save money if you diy but you'll first need to know your average consumption and peak demand

You can use pvgis to estimate how much you need in panels and batteries for example my sweet spot seems to be 7-8kw of panels and 30kwh of batteries

 

[OzSolar]

I may be oversimplifying my calcs, but here they are. Let me know if I should be more conservative or...just totally messed up the math. Seems like most solar calcs must be using the same or similar formulas because my percentages come out almost identical to say...tesla projects or other sites, depending on the size.

(Panels x watts) x .85 x sun hours for location
((12 x 400) x .85) x 5.2 = 21216watts/day
21216 x 365 (yearly projected watts produced) / 1000 = 7743.84 kW annually
7743.84kW x .15 (my average kWh cost)

FWIW: Generally speaking from my experience in the midwest 1,500 kWh (AC) per kW per year is the starting point for an array at the optimal tilt pointing due south and in the sun all day long.

You're at a bit over 1,600 kWh/kWp which might be possible for a perfect site in Colorado.

 

[jlrosine]

Everyone seems to think I should get Emporia. What advantage would that give me over just identifying how much power I use and when that power is used in total kW per day? Is it just a way for you to determine how to segment your home if you decide to use more panels (setup critical load panel) etc.? It seems like a cool thing to have but not really necessary...

 

[1201]

L

Everyone seems to think I should get Emporia. What advantage would that give me over just identifying how much power I use and when that power is used in total kW per day? Is it just a way for you to determine how to segment your home if you decide to use more panels (setup critical load panel) etc.? It seems like a cool thing to have but not really necessary...

I use an iotawatt which is more expensive but they stopped selling it. The Vue is a bargain for what it can do.

How do you know how to size your inverter? Which loads to power? I use on average 1.5kw per hour but I've seen peaks of over 15kw. That tells me how to size my inverter so I don't go too big or too small. I know that 50-60% of my consumption is HVAC so if I just offset that with solar and nothing else I can save a lot. But maybe I can add my fridge, or my septic, or both, and not go over the capability of of my chosen inverter. I know that my dryer uses 5.7kw so if I want to cover that too I'll need a big expensive inverter so maybe I'll just leave that grid tied so I can get by with a smaller inverter.i know that my minimum base load is 275watts so I can use a 200-250 watt grid tied setup and not worry about zero export or backfeeding the grid by mistake

For the $170 or so it's well worth it

 

[zanydroid]

Everyone seems to think I should get Emporia. What advantage would that give me over just identifying how much power I use and when that power is used in total kW per day? Is it just a way for you to determine how to segment your home if you decide to use more panels (setup critical load panel) etc.? It seems like a cool thing to have but not really necessary...

The question should be framed in, does the ROI on the equipment/installation time pay off versus the much better quality and volume of data, and less error prone analysis that comes out of that.

You can directly answer, with low latency, answer many energy consumption questions you might have. How many solar panels, which appliances to get rid of/timeshift.

Using a laborious technique, low resolution technique like looking at the meter manually or grabbing the slowly updating data from the power company is antithetical to rapid learning, such a long feedback loop and high cost of getting the data.

What would be your alternative? Do linear algebra to identify the contributing components? Guess? There are too many free variables to even solve the linear algebra with just a single reading at the meter, unless you do some kind of inference like Sense is doing (which you can't do, you don't have access to that granularity of timeseries data with staring at the meter and models, also Sense is silly vs multiple channel reading).

Emporia is also better than clamp CT in that it can record multiple timeseries in real time without manual work to collect the datapoints. Clamp CT is still needed for ad hoc, one-off measurements and for analysis like inrush current that these energy monitors cannot do.

That said once you get the intuition for where power and energy go, your guesses get a lot better. But how would you get that intuition quickly, without the right tool and workflow?

 

[jlrosine]

The question should be framed in, does the ROI on the equipment/installation time pay off versus the much better quality and volume of data, and less error prone analysis that comes out of that.

You can directly answer, with low latency, answer many energy consumption questions you might have. How many solar panels, which appliances to get rid of/timeshift.

Using a laborious technique, low resolution technique like looking at the meter manually or grabbing the slowly updating data from the power company is antithetical to rapid learning, such a long feedback loop and high cost of getting the data.

What would be your alternative? Do linear algebra to identify the contributing components? Guess? There are too many free variables to even solve the linear algebra with just a single reading at the meter, unless you do some kind of inference like Sense is doing (which you can't do, you don't have access to that granularity of timeseries data with staring at the meter and models, also Sense is silly vs multiple channel reading).

Emporia is also better than clamp CT in that it can record multiple timeseries in real time without manual work to collect the datapoints. Clamp CT is still needed for ad hoc, one-off measurements and for analysis like inrush current that these energy monitors cannot do.

That said once you get the intuition for where power and energy go, your guesses get a lot better. But how would you get that intuition quickly, without the right tool and workflow?

I actually ordered the Emporia tonight, I'll see how it looks tomorrow.

 

[thewilliam]

Consider using the IQ7 series (minor cost savings) if you don't need the IQ8 enhancements. I recently had 14 IQ7+ installed with 400 watt panels. I looked at the IQ8 specs and decided I didn't need its main feature - grid forming. I do wish I had used the IQ7A - the IQ7+'s are clipping for about 4 hours every day. I haven't actually figured it out yet but I think the incremental cost of the 7A would have been worth it for my install (location, azimuth, tilt). FWIW I am hitting my PVWATTS daily production estimates on sunny days - PVWATTS is a good tool. As far as I can tell if you use Enphase microinvertrrs you are stuck with Enphase storage so if you even have a tickling feeling you might add storage in the future, keep that in mind. I think using the Enphase microinverters simplified the install quite a bit and to me at least it comes closer to an engineered solution which I like, at least for my first foray. To be fair, I have seen some very well installed/thought-out/engineered diy solutions here on this forum.

 

[jlrosine]

Consider using the IQ7 series (minor cost savings) if you don't need the IQ8 enhancements. I recently had 14 IQ7+ installed with 400 watt panels. I looked at the IQ8 specs and decided I didn't need its main feature - grid forming. I do wish I had used the IQ7A - the IQ7+'s are clipping for about 4 hours every day. I haven't actually figured it out yet but I think the incremental cost of the 7A would have been worth it for my install (location, azimuth, tilt). FWIW I am hitting my PVWATTS daily production estimates on sunny days - PVWATTS is a good tool. As far as I can tell if you use Enphase microinvertrrs you are stuck with Enphase storage so if you even have a tickling feeling you might add storage in the future, keep that in mind. I think using the Enphase microinverters simplified the install quite a bit and to me at least it comes closer to an engineered solution which I like, at least for my first foray. To be fair, I have seen some very well installed/thought-out/engineered diy solutions here on this forum.

Thanks William.

Even since I started this post, I have changed my mind 100 times, so I haven't settled on any one solution yet. Due to some feedback here from others I am considering a more home grown solution, maybe starting with the MP LV6548 inverter. This would be more flexible for adding different brands of batteries, and also would allow me to have power when the grid is down etc. The downside to this solution I suppose is that I need to relocate my critical loads to a new panel, and not export power back to the grid. I'd still be using the grid as a source of power to the inverter, but have flexibility of using my generator/batteries.

Still thinking and costing through it all. I'm going to monitor my home power usage (time of day) and have a better idea to the size of system, and whether I should just consider batteries right from the start.

 

[1201]

Thanks William.

Even since I started this post, I have changed my mind 100 times, so I haven't settled on any one solution yet. Due to some feedback here from others I am considering a more home grown solution, maybe starting with the MP LV6548 inverter. This would be more flexible for adding different brands of batteries, and also would allow me to have power when the grid is down etc. The downside to this solution I suppose is that I need to relocate my critical loads to a new panel, and not export power back to the grid. I'd still be using the grid as a source of power to the inverter, but have flexibility of using my generator/batteries.

Still thinking and costing through it all. I'm going to monitor my home power usage (time of day) and have a better idea to the size of system, and whether I should just consider batteries right from the start.

Have you played with pvgis? https://re.jrc.ec.europa.eu/pvg_tools/en/

With the lv6548, you need two, and I would think they would power your whole house. You could put in a distribution block before your main panel to power the inverter, feed the inverter to the main panel and use an interlock to make sure the grid cannot power the main panel at the same time as the inverter. That way, you still have power to the inverter but you don't have to move circuits to a new panel

 

[Solarisium]

Thanks William.

Even since I started this post, I have changed my mind 100 times,

Feedback on solar designs on this forum is diverse. Just like asking Google for anything, you will see many suggestions that are absolutely smart, cost effective, long term and solidly engineered, and sometimes contradictory.

But none is tailored to your specific needs. At the end of the day if your design ends up with a long break even (consider you’ll be moving), and little ROI, think again. If your Utility allows grid tie with annual billing and a fair net metering, your excess energy generation in summer will be credited back to you in winter. Something to think about and take advantage of if available.

 

[kevin93]

Consider using the IQ7 series (minor cost savings) if you don't need the IQ8 enhancements. I recently had 14 IQ7+ installed with 400 watt panels. I looked at the IQ8 specs and decided I didn't need its main feature - grid forming. I do wish I had used the IQ7A - the IQ7+'s are clipping for about 4 hours every day. I haven't actually figured it out yet but I think the incremental cost of the 7A would have been worth it for my install (location, azimuth, tilt). FWIW I am hitting my PVWATTS daily production estimates on sunny days - PVWATTS is a good tool. As far as I can tell if you use Enphase microinvertrrs you are stuck with Enphase storage so if you even have a tickling feeling you might add storage in the future, keep that in mind. I think using the Enphase microinverters simplified the install quite a bit and to me at least it comes closer to an engineered solution which I like, at least for my first foray. To be fair, I have seen some very well installed/thought-out/engineered diy solutions here on this forum.

As far as I can tell if you use Enphase microinvertrrs you are stuck with Enphase storage

Not true - any AC coupled batteries can be used.

I have Tesla Powerwalls with Enphase IQ8M-72-2-US inverters.

One limitation is that since the Tesla Energy Gateway cannot control the micro inverters when in grid-tied operation they cannot be configured for zero-export, even in self-powered mode. If the inverters are producing more power than the house plus battery recharge requires the remaining power has to be exported.

In Off-Grid mode the Powerwalls can regulate excess solar production by adjusting the frequency. The inverters follow the Powerwall inverter and adjusting the frequency above 60Hz up to about 62Hz the Enphase inverters drop their production down to zero. (This behavior is defined in UL-1741 - CA Rule 21).

kw

 

[zanydroid]

Not true - any AC coupled batteries can be used.

I have Tesla Powerwalls with Enphase IQ8M-72-2-US inverters.

That is overly optimistic. I think really turnkey/premium AC coupled battery with a large install base have a higher chance of working. There's some megathreads on this forum complaining about IQ8 compatibility with AC coupling to hybrid inverters -- specifically SolArk and Schneider, but I'm sure the Tier2 and lower hybrids would have a worse time at it.

Yes 1741SA (and the interim rules between 1741 and 1741SA) requires frequency-watts adjustment, but that does not guarantee interoperability with whatever you AC couple the microinverters to (Each California POCO or MUD's Rule 21 specifies the precise settings for their local grid).


As another plug for Hoymiles, they have a more modular zero export controller than other manufacturers. There's also a thread talking about how to use this to construct a helper battery that is AC coupled to Enphase batteries.

 

[jlrosine]

As another plug for Hoymiles, they have a more modular zero export controller than other manufacturers. There's also a thread talking about how to use this to construct a helper battery that is AC coupled to Enphase batteries.

If I wanted to go Hoymiles on the cheap (like MPP cheap), is there anything comparable that would allow me to do zero export with CT input monitors on a main existing panel?

Currently I'm debating on just getting a cheaper inverter/smaller setup to get fast ROI and doing net metering......or....going with an offgrid style inverter (or two), and setting up a critical circuit panel and not doing net metering. I'm guessing the first option might yield faster ROI but the second option feels more fun, still need to do the math though.

 

[Shimmy]

I use an iotawatt which is more expensive but they stopped selling it.

Curse you, bearer of bad news!

I love my IoTaWatt's and hope Bob finds peace in his new capacity... and really hope someone else gets UL for the ESP32 version with >32 inputs.

Everyone seems to think I should get Emporia. What advantage would that give me over just identifying how much power I use and when that power is used in total kW per day?

Hopefully you will be amazed at the difference. For me the 5-second IoTaWatt resolution tells me an order of magnitude more than the average daily kWh. Sure, you know what your total load is... but do you understand how things overlap or the profile of the loads? When I couple it with an Influx database, I can correlate temperature and dewpoint to my AC unit operation and know which fan speed makes the system worik for the most comfort.

 

[1201]

If I wanted to go Hoymiles on the cheap (like MPP cheap), is there anything comparable that would allow me to do zero export with CT input monitors on a main existing panel?

Currently I'm debating on just getting a cheaper inverter/smaller setup to get fast ROI and doing net metering......or....going with an offgrid style inverter (or two), and setting up a critical circuit panel and not doing net metering. I'm guessing the first option might yield faster ROI but the second option feels more fun, still need to do the math though.

You can get a $750 growatt inverter , some panels and go grid tied. If necessary the growatt can do zero export. That would be the cheapest. You could try to size it so you consumed most of it and export very little and shift your consumption to the daytime.

In my situation, battery make the most sense because to go grid tied there is an extra monthly charge of $15, and I need an engineer stamped sld and and inspection

Curse you, bearer of bad news!

I love my IoTaWatt's and hope Bob finds peace in his new capacity... and really hope someone else gets UL for the ESP32 version with >32 inputs.

Hopefully you will be amazed at the difference. For me the 5-second IoTaWatt resolution tells me an order of magnitude more than the average daily kWh. Sure, you know what your total load is... but do you understand how things overlap or the profile of the loads? When I couple it with an Influx database, I can correlate temperature and dewpoint to my AC unit operation and know which fan speed makes the system worik for the most comfort.

Lol . I understand your anger ?. I was upset too when I read the bad news.

 

[zanydroid]

It's not clear to me that you have a fully fleshed model of the risk/benefit. Here are the gaps I see:

• Code compliance + home resale liability (important since you are being very precise about what happens after you sell this house in 10 years, yet you say you are going to keep the hardware in place)
• Export / interconnection responsibility (if you use grid tie or hybrid hardware)
• Chance of detecting unsanctioned set up by utility worker driving through your neighborhood / at your house for other business & seeing the panels (off-grid option)
  • There are stories of this in the forum
  • A utility worker comes over to each house around once every 1-2 years to check gas or electric meter

If you are serious about cost savings, in a way that will not leave you to exposure if you sell the place with unpermitted / non-code-compliant hardware. The correct architecture to start exploration from is to start with rapid shutdown devices on a string inverter. You can further cut costs by buying used solar panels (though note you can't take 30% Federal ITC on used equipment, and the racking costs will be higher on smaller panels since those costs are directly proportional to the linear).

You also haven't shared what the roof and shading complexity is of your situation. Roof complexity = how many planes there are, which translates to number of MPPT. Same orientation but split across multiple sections conceivably can go on the same MPPT but have increased racking and wiring costs.

If I wanted to go Hoymiles on the cheap (like MPP cheap), is there anything comparable that would allow me to do zero export with CT input monitors on a main existing panel?

I'm not aware of something cheaper than Hoymiles/other multiport ones that is code compliant in the US.

Hoymiles has a zero export system, it officially requires the DTU and an energy meter compatible with its ModBus interface. There is also cheaper option using OpenDTU and random python script https://github.com/reserve85/HoymilesZeroExport

You are aware that zero export means "zero export on average" and can export a detectable amount, depending on the type of meter that you have? And it counts as parallel operation.

 

[jlrosine]

You are aware that zero export means "zero export on average" and can export a detectable amount, depending on the type of meter that you have? And it counts as parallel operation.

Yeah I think if I do anything (unless it's just a tiny hobby system to learn/play around), I'm going to just pull a permit. $500 for permit + inspection here.

I have a net metering question though, and maybe I just need to call my power coop. It seems that "Net Metering" is always considered as 1 to 1, so if you produce 1kW you get 1kW, depending on the timeframe and usage. However "Net Billing" is the process of the utility company rolling over credits for you until the end of the year, then cutting you a check for unused credits, those are wholesale price.

I guess I'll have to just call the power coop to see, but does anyone have experience with how they perform that net metering vs. net billing? My power company is United Power here in Colorado. https://www.unitedpower.com/solar-bill.

Reading through this it's still not 100% certain, but MAYBE my energy banks are one to one?

"Your net meter energy bank balance will continue to roll over each month until March of each year when all net meter energy banks are settled and returned to a zero balance. All net meter bank balances are paid out at the current wholesale energy rate.

If you have a balance in your net meter bank, you can have your balance applied as a bill credit to your April bill to help offset future service and energy charges"

 

[1201]

It's not clear to me that you have a fully fleshed model of the risk/benefit. Here are the gaps I see:

• Code compliance + home resale liability (important since you are being very precise about what happens after you sell this house in 10 years, yet you say you are going to keep the hardware in place)
• Export / interconnection responsibility (if you use grid tie or hybrid hardware)
• Chance of detecting unsanctioned set up by utility worker driving through your neighborhood / at your house for other business & seeing the panels (off-grid option)
  • There are stories of this in the forum
  • A utility worker comes over to each house around once every 1-2 years to check gas or electric meter

If you are serious about cost savings, in a way that will not leave you to exposure if you sell the place with unpermitted / non-code-compliant hardware. The correct architecture to start exploration from is to start with rapid shutdown devices on a string inverter. You can further cut costs by buying used solar panels (though note you can't take 30% Federal ITC on used equipment, and the racking costs will be higher on smaller panels since those costs are directly proportional to the linear).

You also haven't shared what the roof and shading complexity is of your situation. Roof complexity = how many planes there are, which translates to number of MPPT. Same orientation but split across multiple sections conceivably can go on the same MPPT but have increased racking and wiring costs.


I'm not aware of something cheaper than Hoymiles/other multiport ones that is code compliant in the US.

Hoymiles has a zero export system, it officially requires the DTU and an energy meter compatible with its ModBus interface. There is also cheaper option using OpenDTU and random python script https://github.com/reserve85/HoymilesZeroExport

You are aware that zero export means "zero export on average" and can export a detectable amount, depending on the type of meter that you have? And it counts as parallel operation.

in this day and age with net metering going extinct, Im ready for a true zero export solution to come to the marketplace. with all the technology we have right now that should not be difficult

 

[1201]

Yeah I think if I do anything (unless it's just a tiny hobby system to learn/play around), I'm going to just pull a permit. $500 for permit + inspection here.

I have a net metering question though, and maybe I just need to call my power coop. It seems that "Net Metering" is always considered as 1 to 1, so if you produce 1kW you get 1kW, depending on the timeframe and usage. However "Net Billing" is the process of the utility company rolling over credits for you until the end of the year, then cutting you a check for unused credits, those are wholesale price.

I guess I'll have to just call the power coop to see, but does anyone have experience with how they perform that net metering vs. net billing? My power company is United Power here in Colorado. https://www.unitedpower.com/solar-bill.

Reading through this it's still not 100% certain, but MAYBE my energy banks are one to one?

"Your net meter energy bank balance will continue to roll over each month until March of each year when all net meter energy banks are settled and returned to a zero balance. All net meter bank balances are paid out at the current wholesale energy rate.

If you have a balance in your net meter bank, you can have your balance applied as a bill credit to your April bill to help offset future service and energy charges"

the way Im reading it, its a true 1-1 AND your export even carries over month to month, AND if you have a credit at the end of the yearly cycle they even pay you wholesale rates for any excess? lord I'm so jealous.
consider jumping on this QUICKLY before they change to a "buyback at wholesale" like the rest of the country is RAPIDLY doing.

 

[jlrosine]

the way Im reading it, its a true 1-1 AND your export even carries over month to month, AND if you have a credit at the end of the yearly cycle they even pay you wholesale rates for any excess? lord I'm so jealous.
consider jumping on this QUICKLY before they change to a "buyback at wholesale" like the rest of the country is RAPIDLY doing.

I think this will stay in place for some time, or as long as the current politicians/state leaders stay in place. There was a state law passed in Colorado in 2021 that forced utility companies to do this...they didn't want to but they have no choice right now. Guessing that might change with lobbying etc., we seem to behave like California, just on a big lag. Whatever happens in CA, we usually see it here 5-10 years later...