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

[GregTR]

I just did a post about how my solar system performed over the last 159 days and I calculated my system efficiency which came out to a fairly abysmal 81.3%.

So how did I calculate the efficiency?
I measured my actual load using Emporia Vue inside my main panel. I have compared its readings to the utility meter before I had solar and the measurement was within 0.5-1.0% of the metered reading therefore I consider this measurement fairly accurate and precise.

My Total load side consumption was 10,700 kWh over the 159 day period.

I also got readings from my meter and I consumed 9,274 kWh of grid power during this period and exported 4,984 kWh of solar power to the grid over this same time period.

Therefore my net solar production is what I exported plus what I covered between my load and what I consumed from the grid. This comes out to 6,410 kWh of total Solar production.

However my Sol-Ark data shows that I produced 8,676 kWh and my Tigo optimizers registered 8,871 kWh. This means there is a 2.2% loss over the wires from the panels to the inverter.

If I look at strictly solar then 8,871 kWh on the roof ended up being 6,410 kWh at the load side which is 72% efficiency but I'm not sure that really covers the whole picture.

I'd say there is a 2,461 kWh loss in generating 10,700 kWh of energy for my house which comes out to to be 81.3% or 18.7% loss.

This is quite a surprising number as I would not have guessed that I'm losing this much energy within the system. Granted the battery draws 100W constant and the Sol-Ark has an idle power of 90W I'm still surprised that on average there is 644W of power just getting turned into heat somewhere along the way constantly.

Has anyone else done a similar calculation?

 

[timselectric]

It is what it is.
Hopefully you have a great NEM agreement.

 

[DougfromdaUP]

So you’re losing 644 watts between your inverter and your service? That must be one warm cable. Something sure doesn’t sound right.

 

[GregTR]

It is what it is.
Hopefully you have a great NEM agreement.

I get paid $0.03 for every kWh I send to the grid and I pay $0.29/kWh between 7AM and 9PM and pay $0.00/kWh between 9PM and 7AM. I think this is pretty good

But ROI on my system in Texas is still 18+ years....

 

[robby]

Is this the 12K or 15K model?
I did the math on my 12K not long after installing it and got something like 94% efficiency and the idle power was exactly 60W as advertised.
My advise is to email Sol-Ark your numbers and have them double check your account and show you what they have calculated it to be based on the numbers you supply plus what they are seeing from the Inverter.
I highly doubt the efficiency could be 81.3% when it is CEC listed as 96.5%

 

[Texas-Mark]

I pay $0.29/kWh between 7AM and 9PM

But ROI on my system in Texas is still 18+ years....

Where in TX? I only pay $0.11/kWh which makes going solar not really cost effective.

Note: I do have a small 2.5kW setup, but it is more of a hobby than to try and lower my bill or to be off grid. With all of the upgrades/changes I have done over the years the ROI is long after I am in the grave. LOL

 

[GregTR]

Where in TX? I only pay $0.11/kWh which makes going solar not really cost effective.

Indeed, solar is not cost effective at all in TX. I'm in ONCOR territory and I am with Just Energy and on their free nights plan so daytime they charge a lot but I use no electricity from the grid during the day, only at night.

 

[GregTR]

Is this the 12K or 15K model?
I did the math on my 12K not long after installing it and got something like 94% efficiency and the idle power was exactly 60W as advertised.
My advise is to email Sol-Ark your numbers and have them double check your account and show you what they have calculated it to be based on the numbers you supply plus what they are seeing from the Inverter.
I highly doubt the efficiency could be 81.3% when it is CEC listed as 96.5%

So I think Sol-Ark is doing a bit of arithmetic gymnastics. If you look in their actual reporting you will see that they give you PV, Load, Export, Import, Battery Discharge, and Battery Charge values.

If this is all the source of data you have the numbers should still come out to a net zero, meaning:
(PV + Import + Discharge) - (Load + Export + Charge) = 0 = Sol-Ark Delta

However if you look in the chart below (Sol-Ark Delta) you will see that this is not true, the numbers don't add up to zero. A normal consumer would assume "oh, well whatever is missing is the loss of my inverter" and you would see that it roughly adds up to 100 kWh/month which is about 139W of continuous power which is well within the realm of possibilities if idle power is 90W and the higher the production the higher the loss.

And if you take the total delta of 498.7kWh and consider the 10817.9 kWh of Load then you get 95.6% efficiency.

So let's look at the battery efficiency: If we consider the difference between Charge and Discharge, with my Homegrid setup that would be ~69.4 kWh/month which equates to 84.6% efficiency on the battery alone which manifests itself in an additional 3.1% system loss.

But what if you have secondary sources of measurements for the rest of the data? Well that is when you realize that Sol-Ark lies about every single one of them, and it lies in the most beneficial way for Sol-Ark, it never underestimates Load or Grid Export and never overestimates Grid Import.

While the discrepancy is the smallest on grid export, it is quite substantial on the load side. My Smart Meter Import/Export values are from my utility provider and are calibrated and measured. If I take these two measurement discrepancies between what my utility provider reads and what Sol-Ark reports we find an additional 12 kWh/month of discrepancy which seems minimal but contribute another 0.6% system loss.

But the most blatant discrepancy is on the Load side. I am using an Emporia Vue which was very close to what my utility measured before I installed solar, within 0.5-1% of error, therefore I am highly confident that the Emporia Load measurement is far more accurate than the Sol-Ark one. And here, Sol-Ark is seriously wrong. It overestimates consumption by a whopping 5.4% !

And if we assume that the Emporia Load measurement is the correct one then your system efficiency drops to 95.4%, the battery loss is an additional 3.3% system loss and the Load overestimation is also an additional 5.1% system loss.

If we take the aggregate losses (hidden and presented by Sol-Ark) of 1,459.7 kWh over the 10,266 kWh of Load then your net system efficiency is indeed 87.5%

I guess it's not 81% but it's definitely not 95% either.

 

[Mithril]

I just did a post about how my solar system performed over the last 159 days and I calculated my system efficiency which came out to a fairly abysmal 81.3%.

So how did I calculate the efficiency?
I measured my actual load using Emporia Vue inside my main panel. I have compared its readings to the utility meter before I had solar and the measurement was within 0.5-1.0% of the metered reading therefore I consider this measurement fairly accurate and precise.

My Total load side consumption was 10,700 kWh over the 159 day period.

I also got readings from my meter and I consumed 9,274 kWh of grid power during this period and exported 4,984 kWh of solar power to the grid over this same time period.

Therefore my net solar production is what I exported plus what I covered between my load and what I consumed from the grid. This comes out to 6,410 kWh of total Solar production.

However my Sol-Ark data shows that I produced 8,676 kWh and my Tigo optimizers registered 8,871 kWh. This means there is a 2.2% loss over the wires from the panels to the inverter.

If I look at strictly solar then 8,871 kWh on the roof ended up being 6,410 kWh at the load side which is 72% efficiency but I'm not sure that really covers the whole picture.

I'd say there is a 2,461 kWh loss in generating 10,700 kWh of energy for my house which comes out to to be 81.3% or 18.7% loss.

This is quite a surprising number as I would not have guessed that I'm losing this much energy within the system. Granted the battery draws 100W constant and the Sol-Ark has an idle power of 90W I'm still surprised that on average there is 644W of power just getting turned into heat somewhere along the way constantly.

Has anyone else done a similar calculation?

I'm sorry, your BATTERY draws 100W constant? On the one hand I hope that was mean to be plural, on the other hand even then... ?

 

[GregTR]

I'm sorry, your BATTERY draws 100W constant? On the one hand I hope that was mean to be plural, on the other hand even then... ?

It's a 33.8 kWh Homegrid rack. You can see my battery efficiency is 84% from the math above.

 

[time2roll]

Three cents Stop selling.

 

[GregTR]

Three cents Stop selling.

Why? I can't self consume during the day and I get free electricity at night. My batteries are sitting at 100% charge from midnight until 5PM every day. Then I deplete them during the evening peak and recharge from grid at 9PM.

Once my provider agreement changes where I no longer get free nights I will definitely switch to self consumption!

 

[Mithril]

It's a 33.8 kWh Homegrid rack. You can see my battery efficiency is 84% from the math above.

This is a LifePo4 battery bank? No way in [censored] should it be pulling down 100W constantly (the original statement). Nor should you be seeing 84% between charge and discharge from the battery alone (NOT the charger or inverter part).

84% from "energy at inverter in to battery" to "energy from battery out from inverter" I would buy, as you're paying an efficiency on each direction (so if it's 90% each way thats .9 * .9 for 81%)

 

[GregTR]

This is a LifePo4 battery bank? No way in [censored] should it be pulling down 100W constantly (the original statement). Nor should you be seeing 84% between charge and discharge from the battery alone (NOT the charger or inverter part).

84% from "energy at inverter in to battery" to "energy from battery out from inverter" I would buy, as you're paying an efficiency on each direction (so if it's 90% each way thats .9 * .9 for 81%)

It is a LifePo4 bank with 15s config.

It is pulling 45W when it's considered 100% charged


But as soon as I start drawing from it at 7AM (my free electricity stops so I switch to battery) and it dips below 100% it keeps drawing 100W even after solar recharges it after the sun comes up, all day long:

 

[GregTR]

84% from "energy at inverter in to battery" to "energy from battery out from inverter" I would buy, as you're paying an efficiency on each direction (so if it's 90% each way thats .9 * .9 for 81%)

The 84% is the ratio between power into the battery and power out of the battery as measured by the inverter.

 

[Mithril]

If you are storing and consuming energy you have to actually track the energy flow to and from the battery (with a shunt, or a good clamp meter, etc), if you are simply calculating the whole round trip in one go you're double hitting.

Also when using an inverter (or generator) power factor matters unlike most consumer power plans. If all your loads are resistive then Watts is watts is VA and you can easily compare. But once you throw in any inductive loads, anything with any kind of power supply (LED bulbs even) your power factor goes down which WILL increase your VA compared to watts and how badly that hurts efficiency depends on the inverter(s) and the resulting current demand waveform. The maths of which are beyond me, as most of the time people use PURE power factor relationships (shifted sinewave) which don't account for things like power supplies.

 

[Mithril]

It is a LifePo4 bank with 15s config.

It is pulling 45W when it's considered 100% charged
View attachment 139088

But as soon as I start drawing from it at 7AM (my free electricity stops so I switch to battery) and it dips below 100% it keeps drawing 100W even after solar recharges it after the sun comes up, all day long:
View attachment 139090

This feels really broken to me. Do they have internal heaters? Is the BMS doing balancing (and perhaps failing to get to balance)

What software/sensor is this from?

 

[Mithril]

Wait 15s? Is your inverter correctly configured for a 15s battery? 16s is more common I think. Is there closed loop communication from inverter to battery?

 

[GregTR]

If you are storing and consuming energy you have to actually track the energy flow to and from the battery (with a shunt, or a good clamp meter, etc), if you are simply calculating the whole round trip in one go you're double hitting.

Also when using an inverter (or generator) power factor matters unlike most consumer power plans. If all your loads are resistive then Watts is watts is VA and you can easily compare. But once you throw in any inductive loads, anything with any kind of power supply (LED bulbs even) your power factor goes down which WILL increase your VA compared to watts and how badly that hurts efficiency depends on the inverter(s) and the resulting current demand waveform. The maths of which are beyond me, as most of the time people use PURE power factor relationships (shifted sinewave) which don't account for things like power supplies.

I do have a "good clamp meter" a Fluke DC clamp meter and the current and voltage matches that of what Sol-Ark reports (within orders of magnitude).

I understand what you're saying, but my point remains that the Sol-Ark internal measurement of Load is grossly overestimated which benefits the efficiency number that most would see. It's not until you start comparing the numbers from other, verifiable, sources and the error becomes apparent.

 

[GregTR]

Wait 15s? Is your inverter correctly configured for a 15s battery? 16s is more common I think. Is there closed loop communication from inverter to battery?

It is a closed loop comms system. SOC and other values are reporting correctly.

The data I showed is from solar-assistant that is polling the Sol-Ark on a serial interface.

 

[GregTR]

This feels really broken to me. Do they have internal heaters? Is the BMS doing balancing (and perhaps failing to get to balance)

What software/sensor is this from?

I'm in Texas, it'll hit 79 tomorrow, I doubt it is running heaters...

It is possible that it's a firmware bug, I might reach out to Homegrid and ask them what they think of this scenario ie. why am I dissipating away 50W of power during the day. The 45W idle draw is probably legit, the unit has a nice display and I'm guessing a power supply and electronics that can draw that much.

I have some really nice RGBW permanent Christmas lights around my house and they draw 100W constantly even when they're off so there's that....

 

[timselectric]

I get paid $0.03 for every kWh I send to the grid and I pay $0.29/kWh between 7AM and 9PM and pay $0.00/kWh between 9PM and 7AM. I think this is pretty good

But ROI on my system in Texas is still 18+ years....

Ouch

 

[time2roll]

Why?

Probably below cost. Energy might be free at night, equipment costs money. Just wear on the battery and inverter for nothing.
If the utility can't pay more than 3 cents at peak 4p-9p then just stiff 'em

 

[OffGridInTheCity]

This won't address you're specific situation, but I too track efficiency and it might make you feel better?

Mine situation is much simpler. I record the output from my AIMS inverters in kwhs and divide that by the PV input as reported by the Midnite Classics in kwhs. Over the last 4 years I have an average of 83.1% efficiency from PV (at Midnights) to consumable power (from inverters) This metric covers Charge Controllers -> Battery -> Inverters. In addition, I have at least a 2-3% loss in the wiring from PV arrays to the Midnites.

This puts my overall efficiency at ~80%. So you're not alone in this kind of metric

 

[robby]

I get paid $0.03 for every kWh I send to the grid and I pay $0.29/kWh between 7AM and 9PM and pay $0.00/kWh between 9PM and 7AM. I think this is pretty good

But ROI on my system in Texas is still 18+ years....

Maybe I am missing something here. You use about 67KWh per day and you have 33KWh of batteries.
Why are you not just using the TOU to use PV and Batteries to power the House between 7am and 9pm and then use the Grid to top up the batteries after 9pm. It would seem to me that you could completely zero out your electric bill.
I assume I am missing something?

I still think that rather than making accusations you should just call or email Sol-Ark and have them look at your data. I think they have a group of people that specialize in this area.