• Have you tried out dark mode?! Scroll to the bottom of any page to find a sun or moon icon to turn dark mode on or off!

diy solar

diy solar

My System Efficiency is 81.3%. Prove me wrong or show me your math.

Just an FYI: Solar Assistant Just reports the information it receives from the monitored equipment. So, if there's misinformation. It's coming from your equipment.
 
Just an FYI: Solar Assistant Just reports the information it receives from the monitored equipment. So, if there's misinformation. It's coming from your equipment.
Oh yeah, I'm fully aware of that. It's not Solar-Assistant, it's Sol-Ark that is being overly optimistic. The data between what's in solar assistant and what is on the sol-ark website is largely comparable.
 
What you're missing is that the values in Solar-Assistant are not correct. If I look at my
PV gen at the Tigo on the roof vs. what's in Solar-Assistant, the Solar-Assistant is lower (improves efficiency numbers)
Grid Import at the Smart Utility Meter vs. what's in Solar-Assistant, the Solar-Assistant is lower (improves efficiency numbers)
Load at the Emporia Vue vs. what's in Solar-Assistant, the Solar-Assistant is higher (improves efficiency)
Grid Export at the Smart Utility Meter vs. what's in Solar-Assistant, the Solar-Assistant is higher (improves efficiency)

I trust the Smart Utility Meter and Emporia over my Solar-Assistant (Sol-Ark) readings and they indicate a far worse performance than what the Inverter does.
I would bet money you're comparing Watts with VA, even if it's non obvious. Unless all loads are purely resistive, you simply can't do that.

Let me put this another way, JUST running 120V led lights from the inverter my VA and consumption is roughly DOUBLE what you'd expect from looking at watts. There's a reason why commercial power charges more for bad power factor, it cost MORE money to supply the "same" watts. Consumers not caring about PF and how dirty the demand is is a luxury, and unlikely to stay that way forever.
 
It is the inverter that draws the power due to overhead. I think mine draws oner a hundred Watts an hour or 1.2 kWh per day.
That's not what OP said. They claimed the batteries alone are drawing 45-100W.

Regardless, OPs first step needs to be properly identifying these loads and accounting for them. And if needed fixing them. Trying to evaluate system efficiency with unknown loads is silly.
 
I would bet money you're comparing Watts with VA, even if it's non obvious. Unless all loads are purely resistive, you simply can't do that.

Let me put this another way, JUST running 120V led lights from the inverter my VA and consumption is roughly DOUBLE what you'd expect from looking at watts. There's a reason why commercial power charges more for bad power factor, it cost MORE money to supply the "same" watts. Consumers not caring about PF and how dirty the demand is is a luxury, and unlikely to stay that way forever.
That's not what OP said. They claimed the batteries alone are drawing 45-100W.

Regardless, OPs first step needs to be properly identifying these loads and accounting for them. And if needed fixing them. Trying to evaluate system efficiency with unknown loads is silly.

I'm saying the Sol-Ark measured Load does not match the Load measured by Emporia VUE which incidentally is very close to what my utility meter has measured historically before I installed solar. Additionally the Sol-Ark measured grid import/export does not match what my utility measures. Every one of these measurements from Sol-Ark are more beneficial to Sol-Ark's efficiency rating.

Are you saying Sol-Ark is measuring VA while my utility meter is measuring watts? Are you also saying that I'm essentially getting a discount from my utility company hence their measure is lower than the Sol-Ark measurement?

Even if I spot you the above, which I don't, there is a VAST difference between discrepancy of the grid side and the load side measurements by Sol-Ark and before I had PV my Emporia and the utility meter were within 1% of each other at all times. If Sol-Ark is using the same VA measurement on both sides and my utility and Emporia uses Watts then I would assume the error on both the grid side and on the load side would be similar but it isn't, it is much worse on the load side.

I don't need to identify loads and account for them. I'm only pointing out that based on what the utility meters at my house and what I'm consuming the PV efficiency is below 85% while Sol-Ark's data would indicate a much more optimistic number. The only number I really care about is what my utility is charging me, somehow I doubt they're in the business of losing money...
 
The batteries alone do indeed draw that. And on days like today where I have constant cloud cover movement the draw fluctuates far more during the day as the Sol-Ark is sinking micro bursts into the battery or does micro draws as it's adjusting the grid export on the AC side. You can see that it goes as high as 400W at times and it draws as much as 100W at times. Looking at the chart it's certainly dumping more into the battery than it's drawing from it which will manifest itself as a energy loss.

But this loss is reported and understood in my math, it's not insignificant but certainly doesn't cover the majority of the discrepancy.

battery_fluctuations.jpg
 
What you're missing is that the values in Solar-Assistant are not correct. If I look at my
PV gen at the Tigo on the roof vs. what's in Solar-Assistant, the Solar-Assistant is lower (improves efficiency numbers)
Grid Import at the Smart Utility Meter vs. what's in Solar-Assistant, the Solar-Assistant is lower (improves efficiency numbers)
Load at the Emporia Vue vs. what's in Solar-Assistant, the Solar-Assistant is higher (improves efficiency)
Grid Export at the Smart Utility Meter vs. what's in Solar-Assistant, the Solar-Assistant is higher (improves efficiency)

I trust the Smart Utility Meter and Emporia over my Solar-Assistant (Sol-Ark) readings and they indicate a far worse performance than what the Inverter does.
Interesting. I checked it out like you said, and the Tigo is reporting about 3.8% more PV than Sol-Ark. Thats crazy! I wonder if you could bring this up to Sol-Ark if there is maybe a calibration that could be done?
The batteries alone do indeed draw that. And on days like today where I have constant cloud cover movement the draw fluctuates far more during the day as the Sol-Ark is sinking micro bursts into the battery or does micro draws as it's adjusting the grid export on the AC side. You can see that it goes as high as 400W at times and it draws as much as 100W at times. Looking at the chart it's certainly dumping more into the battery than it's drawing from it which will manifest itself as a energy loss.

But this loss is reported and understood in my math, it's not insignificant but certainly doesn't cover the majority of the discrepancy.
This graph you show is interesting because I do not have that on my end. I might be wrong, but did I read somewhere you are purely AC coupled? I wonder if that has something to do with your wild swings? Heres mine from today once the battery reached 100%.
Battery.jpg
 
W vs. VA could even be different measuring points and what SolArk does with power it receives.
What if external meters measure actual watts when driving an inductive load (part of the power is returned out of phase each cycle),
and what if SolArk burns off that power rather than returning it to capacitors?
Other piece of the puzzle would be how SolArk measures/reports.

A couple test loads (with nothing else powered) could clarify that issue.
A space heater is resistive load. A transformer like the 25kvA one I got and tested is inductive, and draws about 1000W apparent power.
 
Interesting. I checked it out like you said, and the Tigo is reporting about 3.8% more PV than Sol-Ark. Thats crazy! I wonder if you could bring this up to Sol-Ark if there is maybe a calibration that could be done?

This graph you show is interesting because I do not have that on my end. I might be wrong, but did I read somewhere you are purely AC coupled? I wonder if that has something to do with your wild swings? Heres mine from today once the battery reached 100%.
I have about 2.2% loss between the Tigo and the Sol-Ark reported PV which is expected as that is likely lost on the 10awg wire from the roof to the inverter. I'm actually happy that it's under 3%.

What battery do you have? It could be a Homegrid specific issue on the battery side, but that is only "some" of my efficiency loss. I'm an DC coupled, I have nothing on the AC side other than loads.

Here is a day with full sun and no cloud fluctuations. Still a constant draw of ~100W and 97% SOC which never budges, over hours on end.
battery_smooth_PV.jpg
 
A couple test loads (with nothing else powered) could clarify that issue.
A space heater is resistive load. A transformer like the 25kvA one I got and tested is inductive, and draws about 1000W apparent power.
I could definitely do that but my house has a base load of about 1kW so turning everything off just to do a test would be painful and not sure if the juice is worth the squeeze.
 
Leave that on and add 2.0 to 3.6kW of electric space heater? See if incremental measured watts is correct?

What does your base load consist of?
Have a way to see current waveform? PF, real/apparent power of the entire load?

Can SolArk be bypassed so house powered by grid, allowing it to be tested separately?
I have interlocked breakers. Some things ride through, some could be glitched.
 
Leave that on and add 2.0 to 3.6kW of electric space heater? See if incremental measured watts is correct?

What does your base load consist of?
Have a way to see current waveform? PF, real/apparent power of the entire load?

Can SolArk be bypassed so house powered by grid, allowing it to be tested separately?
I have interlocked breakers. Some things ride through, some could be glitched.
My base is LED lights, several computer power supplies/switches, a pool pump and a hot water circulating pump and a TON of smart switches (z-wave, lutron, zigbee)

So I haven't run the test directly but I did look at historical data where I actually run space heaters at night until 7AM (free electricity at night) Unfortunately I only have 15 minute aggregate data this far back from Emporia but the numbers should be still matching.

resistive_estimate.png

You can see that on the space heaters the difference is 4.5% between what Emporia measures and what Sol-Ark measures. The difference on the base load is over 8%.

At 7:55 my wife turned on the dryer so the last line has my dryer's resistive load in it as well as its motor load, you can see that the delta is much closer there, it's down to 3%.

So there is certainly truth to the theory of VA vs. W here but even on purely resistive loads the Sol-Ark is still way off from what Emporia measures which is in line with what my utility company measures.
 
LED and SMPS, that's what I was thinking of. If not PF corrected, the clip the top off sine waves. No power shoved back out of phase (despite their capacitors), but spikes of current cause excess losses in inverter transistors. It is enough to trip my thermal breakers feeding poll pump VFD (an SMPS)

Your pumps may have out of phase current.

Have a scope to look at waveforms? I think charging capacitors at peaks of sine wave, brief spikes of current, will be considerably higher losses (compared to same power consumed by resistive load).

Look at 3rd plot, green trace. That's the AC current.


Ideally one would have PF corrected SMPS, which synthesize a sine wave current on input. I think new Dell have that.
 
I believe the efficiency losses are more towards from the panels to the SCC since the distance of the wire is at the greatest. Batteries do self discharge just sitting on the shelf and each extra fuses and disconnects add more to your loses. The only problem that I see is the amount of money was spent to generate the electricity. That’s where it became inefficient.. only if we can capture the heat off our forehead !!
 
Ya know 20 minutes on the phone with a Sol-Ark power specialist would probably clear this right up.
But like the "My Sol-Ark 15K will only produce 13KW thread" we will go on here for weeks or months until the OP finds out that something else is not right.
 
Ya know 20 minutes on the phone with a Sol-Ark power specialist would probably clear this right up.
But like the "My Sol-Ark 15K will only produce 13KW thread" we will go on here for weeks or months until the OP finds out that something else is not right.
I doubt it...

There is really nothing they can say that would validate why the values I get from Sol-Ark do not match what I get from my utility company. And even if they did explain, ultimately it's the utility company's numbers that I have to abide by and pay for.

But just to make you feel better I'll fire off an email tomorrow to them to see how they can explain the discrepancy.

Just for the record, I'm still liking my Sol-Ark and I still recommend it to everyone I know, I'm just not a fanboy who can't see that their data is wrong and wrong in the direction of improved efficiency.
 
And just to clarify there are several issues at hand here.

1) Battery efficiency. This is likely a Homegrid discussion, I'm confident that Sol-Ark has nothing to do with why the battery is drawing 100W when it's close to full but not quite full yet it never becomes full.

2) Discrepancy in reported grid import and export. This is not as much of a difference as the other items but still significant. Here should be no difference.

3) Discrepancy in reported load measurement. While some of it can be explained with VA vs. W measurements, there is still a 5% error even with resistive loads when compared to Emporia. I trust the Emporia numbers as they are matching that of my utility meter. I doubt my utility would be happy with a meter that is under-reporting by 5% on resistive loads even.

Number 2 and 3 are strictly in Sol-Ark's hands and I doubt thy can give a reasonable explanation, but we shall see.
 
I looked up data sheets for SolArk and Homegrid. I can see why you would expect something better than 90% overall.
There wasn't a curve vs. load, just numbers quoted at 65% load and the 90W no-load consumption.
90W + 3.3% of 7.8kW would be the 95.5% battery to AC efficiency. I expect the 3.3% to go up/down based on square of current.

Idle power could explain up to 2.6% for your total consumption.

The scope traces of current into VFD (rectifier-capacitor) that I linked carried current about 1/3 of the time. If the current was 2x or 3x higher than for same watts delivered to a resistive load, losses would have been 4x (13.2%) or 9x (29.7%), so 9.9% or 26.4% greater than anticipated. The 9.9% greater is around what you're reporting.

If course, if your loads were less than 65% of full load, that same current squared factor would mean lower losses. (in the limit, just leaving 90W idle consumption, 343 kWh during your time period.)

Could be the push for energy efficient LED bulbs is causing you to waste energy.

Pool pump - I did look up specs for variable speed pumps a while back, and they had good power factor correction.

"Discrepancy in reported load measurement" ... "5% error even with resistive loads"

Were you able to separate out resistive loads? Those, I expect inverter to meet specs.
Don't know where it measures delivered power. If it measured average DC current and voltage, that would be power into inverter rather than out of it.

I'm inclined to think losses in FETs and inductors are higher due to PF not 1.0
 
Last edited:
Were you able to separate out resistive loads? Those, I expect inverter to meet specs.
I did by subtracting the base from the power in those equations above. I was still at 4.5% difference between what the Sol-Ark measured and what my Emporia/smart meter measured.
 
Leaving 95.5%, which is about the battery to AC spec (at 65% load)
Maybe SolArk measures voltage and current feeding inverter, rather than having sensors to measure AC power? That would not include 48V battery to HV DC losses.

Do you have a clamp ammerter? That could measure AC RMS current (cheap ones less likely to be true RMS.) AC VA / Vrms = Irms. Watts may differ. But compare clamp measured Irms; if waveform isn't sine wave, has shorter higher spikes, that would have more I^2R loss in FETs and inductors compared to a nice "real" resistive load. Such a meter reading migh explain your results.
 
Leaving 95.5%, which is about the battery to AC spec (at 65% load)
Maybe SolArk measures voltage and current feeding inverter, rather than having sensors to measure AC power? That would not include 48V battery to HV DC losses.

Do you have a clamp ammerter? That could measure AC RMS current (cheap ones less likely to be true RMS.) AC VA / Vrms = Irms. Watts may differ. But compare clamp measured Irms; if waveform isn't sine wave, has shorter higher spikes, that would have more I^2R loss in FETs and inductors compared to a nice "real" resistive load. Such a meter reading migh explain your results.
Is Emporia revenue grade?
 

diy solar

diy solar
Back
Top