diy solar

diy solar

EG4 18k PV Discharging 10 - 50 times per day

I'm curious about the "discharging batteries" requirement.

My 18Kpv does exactly what you're saying above except I don't consider it "discharging" the battery, it just consumes power as needed from the battery. I guess that does, in fact, discharge the battery but not as a forced thing, but as an as-needed thing. I have disabled the "Forced Discharge" setting and this still gives me the behavior I desire.

I am sure that EG4 folks and everyone else understands this better than I do. Can you please tell me what you mean about discharging? Are you trying to force a discharge or just want the inverter to use the batteries at night? If it's the latter, my situation is working exactly like that.
the goal is for the battery to discharge completely overnight...so that the battery can be re-filled the following day.
of course if only 1 kwh is used overnight on a 15kwh battery, then the battery will only discharge 1 kwh.
after 20:00, all power should be coming from the battery first, GRID (bypass) second.
It doesn't have to be a forced discharge... as-needed is fine.

If this is what your 18k pv is doing, what settings are you using to accomplish? Grid-shaving, battery back-up, anything in particular?
 
OK, I see how we're different. I want to use my batteries instead of the grid - all the time. My main concern is having power at all times - and our grid goes down far too often. If I drained my batteries over night it would be just my luck that the power would go out first thing in the morning due to severe weather that lasts all day! We have enough batteries to last us a few days so, theoretically, it should be quite difficult to drain them down to zero - especially over night.

To your question of settings, I am using the system in "self consumption" mode. On the web monitoring, the "maintenance" tab has a "working mode" button. When I click that, I select self consumption and these are the settings I use:

General: Normal, Off Grid Mode - disabled
EPS Output: Power Backup: Enabled (I have a generator attached)
Protection Setting: RSD Enabled (not pertinent to this topic, I realize)
Grid Sell: Grid Sellback enabled @12kw, Fast zero export disabled
Battery Backup Mode: disabled
Battery Setting: Charging - SOC control at 101% with charge last disabled
Discharging: Image attached since this is the crux of your interest

Though it doesn't show it on the working mode screen, the main maintenance screen shows that my "forced discharge" is disabled.

With my configuration, we use batteries until the PV produces enough to run the house. The batteries recharge once the PV input exceeds the demands of our home. The only time we use the grid is when our batteries get lower than 45%. If the grid goes down and can't recharge, then my generator will kick in at 20% and pull the batteries up a bit until the sun comes up.

This gives us an energy usage priority of: PV, Batteries, Grid, then Generator.

I realize everyone can have their own vision and comfort values. This is just how we decided to make use of our system.

Please let us know if you get things working like you expect, and how you did it.
 

Attachments

  • DischargeSettings.png
    DischargeSettings.png
    36.1 KB · Views: 4
OK, I see how we're different. I want to use my batteries instead of the grid - all the time. My main concern is having power at all times - and our grid goes down far too often. If I drained my batteries over night it would be just my luck that the power would go out first thing in the morning due to severe weather that lasts all day!
You are on a design that favors backup power over ROI. This is common amongst off-grid architectures and households with unstabile grids.

The solar industry has catered to your situation for decades. The problem for the solar industry is it wants to become mainstream and in mainstream America we don't have unstabile grids, we aren't off-grid and we want design to favor ROI. Any new entrants to this inverter market should understand this VERY BASIC need, but apparently they are still designing software to favor situations like yours, where backup power is the priority.

Under my desired scenario, the battery is banking unused pv power during the day, and discharging that pv power at night. VERY BASIC.
Grid sell-back rates aren't profitable, so people are buying battery banks to store their pv power.

Thanks for your settings. It appears the EG4 cooperates in situations like yours that are more off-grid style (in terms of how you leverage your battery)...but I'm trying to use the battery for ROI and EG4's software just isn't taking that into consideration.
 
Here's another aspect of how the 18k pv doesn't optimize for ROI. I should have 10kw coming from panels at 12:35pm EST, but I'm only getting 5kw because the inverter is only supplying the needed load, and has already filled the battery early in the day with pv power that wasn't in excess of loads.

Further explanation:

If the 18k pv prioritizes all pv power to battery first, then the batteries get prematurely filled, and then later in the day when pv power is sure to exceed load, there is no battery capacity left to bank the excess pv power. So then pv power is rate limited by the inverter.
Yes, I could sell back to grid at a few pennies, but I bought expensive batteries to bank the pv power, not sell back to grid for pennies.


View attachment 220395
 
the 18k pv doesn't seem to be capable of this priority:
1. powering loads from pv first
2. charging battery with any excess pv
3. power from grid (not battery) when pv power is insufficient
4.. only discharging batteries overnight

I kind of have this setup.

I use the setting “pv charge priority” to completely prevent battery discharge at certain times during the day. So if you set it enabled and set the time for daytime hours say 7am to 7pm the battery will not discharge at all during those times. But yes it then takes priority over point #1. So adjust the times to say 9am and set charge last to enabled then from 7am to 9am you have 1 and 2 sorted. Then at 9am pv charge priority kicks in and switches 1 and 2 around. I also limit the pv charge priority charge rate so as to gently charge the batteries through the day. Then at 7pm or whatever time when pv charge priority ends, battery power is then available as the source. So currently my last time is 6pm to 10pm so that by 10pm all major appliances like oven, stove, ac, hot water have been used for the day and it’s just the normal lights, tv, fridges etc and they all then just use my battery throughout the night till the next morning. So I basically use my full battery capacity everyday.
 
It's interesting how we can all see things so differently. :)

For me, there's no such thing as having the batteries prematurely charged. The earlier my batteries charge up, the sooner we make use of heavy loads like laundry, canning, dinner prep, etc. We like to use as much as we can during the day - once the batteries are charged.

We do export excess power to the grid to run our meter backwards. For now, our POCO has a 1:1 metering arrangement. We didn't get into this for the utility bill but once everything was installed, I felt like I should no longer pay an electric bill. I laugh at myself for the mental transformation that I underwent after installing our system.
 
Here's another aspect of how the 18k pv doesn't optimize for ROI. … It appears the EG4 cooperates in situations like yours that are more off-grid style (in terms of how you leverage your battery)

Isn’t it difficult to write software that can optimize for all ROI scenarios?

For instance under NEM2 in California you should do forced export.

Under NEM3 new software is needed, to do forced export during specific hours in August and September where export pricing is 50x higher.

The 18kpv looks decent enough for me to configure for ROI optimization on NEM2. In the summer I can forced export at 10kW from 7-10 for $.65/kWh and pull it back in winter for $.20/kWh

I don’t believe 18kpv have NEM3 harvesting logic. Enphase is building it soon for their batteries. It is probably fine since my POCO hasn’t implemented the billing yet for NEM3, but will be in place next year.
 
Last edited:
It's interesting how we can all see things so differently. :)

For me, there's no such thing as having the batteries prematurely charged. The earlier my batteries charge up, the sooner we make use of heavy loads like laundry, canning, dinner prep, etc. We like to use as much as we can during the day - once the batteries are charged.

We do export excess power to the grid to run our meter backwards. For now, our POCO has a 1:1 metering arrangement. We didn't get into this for the utility bill but once everything was installed, I felt like I should no longer pay an electric bill. I laugh at myself for the mental transformation that I underwent after installing our system.

Yes I also use the battery during the day but I generate so much pv during the day that is covers 99% of the loads by itself and then just a little battery now and then. I definitely want to use at least a full battery cycle once a day. I just prefer gentle use of the battery both in its discharge and charging phases. I also sell back to the grid so maximizing my roi and my bill has been negative for the past few months. A good example was today, I imported 3kwh, generated 100kwh, consumed 36kwh and exported 60. I also see all my appliances as watts getting used now. Weird. 😃
 
Isn’t it difficult to write software that can optimize for all ROI scenarios?

For instance under NEM2 in California you should do forced export.

Under NEM3 new software is needed, to do forced export during specific hours in August and September where export pricing is 50x higher.

The 18kpv looks decent enough for me to configure for ROI optimization on NEM2

Way to many scenarios so the end user, that would be us, would have millions of options in setting it up. I for example would like to see not set times as the sun apparently doesn’t own a watch. I would like to see an option on sunrise, sunrise + x hours, sunset, sunset + x hours as options for all these settings. Heck we don’t even have TOU rates here, it’s all straightforward rates plain and simple. 0.13 to get 1kwh and 0.06 to export 1kwh everyday all day all year long.
 
Way to many scenarios so the end user, that would be us, would have millions of options in setting it up. I for example would like to see not set times as the sun apparently doesn’t own a watch. I would like to see an option on sunrise, sunrise + x hours, sunset, sunset + x hours as options for all these settings. Heck we don’t even have TOU rates here, it’s all straightforward rates plain and simple. 0.13 to get 1kwh and 0.06 to export 1kwh everyday all day all year long.
Yeah it’s not exactly what I would want but it gets close enough. There are people here that use an external controller to reprogram it.

Enphase and Tesla take the approach of coming up with more fixed programming that is POCO compliant and works decently OK for a lot of pleb customers.
 
I kind of have this setup.

I use the setting “pv charge priority” to completely prevent battery discharge at certain times during the day. So if you set it enabled and set the time for daytime hours say 7am to 7pm the battery will not discharge at all during those times. But yes it then takes priority over point #1.
Yes, you acknowledge that this is not the desired scenario. The problem with your setup is that its running a FIXED schedule. The desire is to have dynamic EXCESS PV feed into battery. This can add up to hundreds of dollars per year difference for a 15kwh battery and thousands per year for a 60 kwh battery because if I'm filling up my battery on a fixed schedule early in the day...with energy that could have gone directly to loads...and then later in the day when the battery is full..my inverter is rate limiting pv production because it has nowhere else to put it. It has to be a dynamic EXCESS PV to battery scenario, no FIXED schedules, or over time, ROI leakage becomes significant.
 
Isn’t it difficult to write software that can optimize for all ROI scenarios?

For instance under NEM2 in California you should do forced export.

Under NEM3 new software is needed, to do forced export during specific hours in August and September where export pricing is 50x higher.

The 18kpv looks decent enough for me to configure for ROI optimization on NEM2. In the summer I can forced export at 10kW from 7-10 for $.65/kWh and pull it back in winter for $.20/kWh

I don’t believe 18kpv have NEM3 harvesting logic. Enphase is building it soon for their batteries. It is probably fine since my POCO hasn’t implemented the billing yet for NEM3, but will be in place next year.
The software is already over-engineered to accomodate 50 marginal use cases (including peak shaving)...my scenario is far more basic and universal ...and should be an out-of-the-box functionality.

It's a simple EXCESS PV to BATTERY priority. And Solark 15k does it, apparently (although I'll believe it when I see it).

If Solark 15k does do it, I'm understanding why so many people champion Solark 15k now. I never thought the EG4 18k pv would miss such an obvious PV prioritization. In fact, every technician I talk to thinks the 18k pv does perform this operation...BUT IT DOESN'T.

And worse, I have problems just getting the EG4 18kpv to discharge batteries...with or without Forced Discharge...which is a WHOLE NOTHER LEVEL of glitching.
 
I don’t know how EG4 prioritized features

FWIW peak shaving would make some sense given that there are probably a couple million US residents that have a POCO with demand charges.
 
A good example was today, I imported 3kwh, generated 100kwh, consumed 36kwh and exported 60. I also see all my appliances as watts getting used now. Weird. 😃
Yes, when you have lucrative sell-back agreements, it leaves you far more options in capitlizing on ROI. For the vast majority who don't have lucrative sell-back agreements, your cited example would be a disaster in ROI (on a daily basis). If EG4 wants to make it in this business, they have to design for my use case which is the vast majority of people who don't have lucrative sell-back agreements.
 
I don’t know how EG4 prioritized features

FWIW peak shaving would make some sense given that there are probably a couple million US residents that have a POCO with demand charges.
yup, peak shaving is a decent size mkt but it pales to the market of households with no 1:1 net metering agreement who must be extremely efficient with how kwh are banked throughout the day....dynamic EXCESS PV to BATTERY.
 
Yes, you acknowledge that this is not the desired scenario. The problem with your setup is that its running a FIXED schedule. The desire is to have dynamic EXCESS PV feed into battery. This can add up to hundreds of dollars per year difference for a 15kwh battery and thousands per year for a 60 kwh battery because if I'm filling up my battery on a fixed schedule early in the day...with energy that could have gone directly to loads...and then later in the day when the battery is full..my inverter is rate limiting pv production because it has nowhere else to put it. It has to be a dynamic EXCESS PV to battery scenario, no FIXED schedules, or over time, ROI leakage becomes significant.
Yep. The charge last is the one I use so that loads are powered in the morning first with any excess going to battery to my charge limit and then export. But as I clip from around 11am that’s also the time I want my battery to charge so more of my pv can be used and I would generate more pv. But my battery might already be fully charged by then. So I need more batteries to capture that “wasted” power when clipping, but to save 3kwh of offset in import doesn’t justify getting any more batteries as that additional batteries would never roi. Yep as mentioned I would like sunrise and sunset as options as well so I don’t use a set time. And then it’s a whole different story if it’s not a normal bright sunny day. And weather forecasting doesn’t help. It was supposed to be a partly cloudy day here today….

So we make suggestions to eg4 of what we need and hope they update their software so we can do what we want it to do. And we make the best of the options we currently have and try to optimize it. It’s always going to be a compromise somewhere for someone. So you can do 1 and 2 very easily just with last charge only. It’s point 3 that breaks it combined with point 4. Seems you can then have 3 or 4 and not both without you changing settings twice a day. You basically want the option to not use battery for loads during the daylight hours while the grid is up. But wouldn’t this then decrease your roi by not using your batteries to power your shortfall on pv during the day? I try and use the heck out of mine as much as possible. Do you use more power at night than during the day that you want to not use grid at night? I guess it’s very dependent on your situation, your loads, your grid provider charges, your battery size and how much pv you have and what the weather is doing. Anyway, I was rambling and digressing. Doesn’t seem like too big an issue to be coded along with sunrise/sunset options. We can all hope 🙂
 
Another thing that has an affect on ROI is battery heat. So I keep this in mind when trying to program my 18Kpv. I try to charge and discharge at a lower amp rate to reduce the heat. And in the cooler part of the day. The chart below gives an example of the degradation that heat plays on the cycle of lithium batteries.
 

Attachments

  • IMG_5546.png
    IMG_5546.png
    115.4 KB · Views: 12
Another thing that has an affect on ROI is battery heat. So I keep this in mind when trying to program my 18Kpv. I try to charge and discharge at a lower amp rate to reduce the heat. And in the cooler part of the day. The chart below gives an example of the degradation that heat plays on the cycle of lithium batteries.
Yes, with batteries being so expensive, every care must be taken to preserve the life. Charge/discharge rate matter which is why being able to charge batteries with excess pv is the most appropriate manner to charge batteries (spare the fixed schedule brutal charge in favor of the casual re-charge over a day or many days.) If load matches pv power on the hour, daily, for a week, you'd go the entire week without charging or discharging your batteries once (savings 7 cycles). That's a 7x battery life increaser (in this scenario). You could tweak these setttings, if you favor having battery backup in case of grid outage, to keep a minimum soc on the battery, but again, ROI is king for most people, not having battery backup on the extremely rare occasion of grid failure.
 
Last edited:
Yes, with batteries being so expensive, every care must be taken to preserve the life. Charge/discharge rate matter which is why being able to charge batteries with excess pv is the most appropriate manner to charge batteries (spare the fixed schedule brutal charge in favor of the casual re-charge over a day or many days.) If load matches pv power on the hour, daily, for a week, you'd go the entire week without charging or discharging your batteries once (savings 7 cycles). That's a 7x battery life increaser (in this scenario). You could tweak these setttings, if you favor having battery backup in case of grid outage, to keep a minimum soc on the battery, but again, ROI is king for most people, not having battery backup on the extremely rare occasion of grid failure.
Apparently Luxpower is doing all the firmware. EG4 is just trying to make the firmware work and that seems to be a problem.
So any changes must come from Luxpower.
This is evident via the recent 18kpv thread where two inverters fried some appliances by applying 350+ volts to the input terminals.
 
Apparently Luxpower is doing all the firmware. EG4 is just trying to make the firmware work and that seems to be a problem.
So any changes must come from Luxpower.
This is evident via the recent 18kpv thread where two inverters fried some appliances by applying 350+ volts to the input terminals.
Give it a break, it's obvious that you have a personal grudge against EG4 and it's pathetic.
 
If EG4 wants to make it in this business, they have to design for my use case which is the vast majority of people who don't have lucrative sell-back agreements.
You have identified a valid improvement that could be made in the programming, maybe give them a touch more than a week to implement it.
 
Give it a break, it's obvious that you have a personal grudge against EG4 and it's pathetic.
No, just reading and observing.
Are you paid by SS or Eg4?? You are constantly coming to their defense when there are obvious issues with this product! I mean, give it a break! I think SS had made some big changes to try and fix their customer service. Its showing. Still, this product has a boatload of issues. And that seems to go right back to Luxpower.
 
You have identified a valid improvement that could be made in the programming, maybe give them a touch more than a week to implement it.
I'd give them a few months if I had an honest back and forth conversation with them. But i'm getting denial vibes and getting piece-meal troubleshooting. I'm coming at this from an installer point of view and I don't want to have to be making excuses for this glitch behavior. Way too much hassle to put this system in client homes.
 

diy solar

diy solar
Back
Top