trouble with enphase envoy consumption measurements

[brianbsl]

Just wanted to follow up on my own issue for anyone who finds the thread in the future - after working with Enphase support to take photos of all of the CT's, I found my installer had broken a clip on one of the CT's when they installed and just kinda jammed it together. But without it being "tight" it was not inducting the full current into the CT. I had the installer come out and replace it and now my data is "good", as in comparable to my Fluke clamp on meter and my Sense.

[brianbsl]

Sorry to dig up a couple month old thread, but did you ever find a solution? I have noticed that my consumption metering is about 15% off compared to my net meter adjusted by my generation meter (our new state program uses two utility provided meters). My production seems to be on-target. I also have a Sense (only on one of my panels but I can compare some large loads like EV charger) and the real-time numbers from the sense seem to line up with the utility meter and not the Enphase meter.

I can't use toolkit because it was not a DIY install but I did pay the extra $ for ElightenManager. It took my installer 4 attempts to get the consumption monitoring working even remotely correctly, so I think there is zero chance I will get them to dive into this problem. I did open a ticket with Enphase today and they told me an engineer will look into it and get back to me in 2 days, but based on what you posted, it seems unlikely that I will get anything back.

Kinda frustrating since the major reason I went with Enphase over SolarEdge was that Enphase officially supported doubling up consumption CT's on my dual 200A panels whereas that was unclear from SolarEdge (although maybe would have worked fine since they support a 400A set of CT's, which I guess could have been installed in my meter base, but the rest of the work we did did not otherwise require opening up the meter base by the installer). (My installer didn't charge any premium for Enphase either as they prefer it).

[ART005]

Last one. Here is heavy overcast at the end of the day with sun low and far to the side of the array. The darkish overcast looked like possible little snow storm coming.
Screenshot_2019-01-06-16-00-21.png

[ART005]

Last one. Here is heavy overcast at the end of the day with sun low and far to the side of the array. The darkish overcast looked like possible little snow storm coming.

[ART005]

Here is low light at the end of the day. This is just low sun with overcast far to the side of the array. Screenshot_2019-01-05-15-46-37.png

[ART005]

Got your message a little before 11am my time so low production is just winding down. Today is medium overcast so that is adding to low production. With all of the typical tight connections I don't have a way to measure volts and compare to expected IQ6+ performance.
Up until 11am I produced 0.4 kWh with my 2.56kwp array.

. Screenshot_2019-01-05-10-53-16.png

[Abe317]

ART005,

ould you share a screenshot of your toolkit during non-production hours showing the production meter reading details, like mine:

https://www.solarpaneltalk.com/filedata/fetch?id=389707&d=1546658760

I am looking to see how our systems' readings compare.

Thanks.

[ART005]

I hope you get the help you need. I am quite sure my problem is based in the start-up voltage and minimum voltage. My 8x320W=2.56 kw mono array gets the IQ6+ inverters running at pf=1 at 700 VA. I can even see it go lower after that and work much better than it did on the way up. While my production meter struggles at PF=0.5-0.7 from 8:30 to 11am in this equinox season, it finally gets to pf=1 as the sun clears the trees with VA=700. I'm producing 1-5 kWh/day and I see almost the exact daily amount reducing my consistent 7 kwh daily usage. Long explanation that I don't think I have any unique problems, just that the mppt curve does not track well below 28% of peak array power. I haven't reached out to Enphase yet. I wonder if they can tweek low light efficiency for me?

[Abe317]

My Enphase IQ6+ based system is a few days old. I'm 43 degrees north latitude with a big hill to the south. About 1 month +/- the winter solstice the sun rises and sets a little more than 1 hr later/earlier than sun charts predict for my location. So I have a lot of indirect sun during that time. From about 8-11am my PF is 0.5-0.7 while a 2.5kWp array puts out ~300VA and ~180W. I had anticipated the 300VA, I'm sure it's correct. But I never realized I would get docked again by PF in the conversion to Grid power. I don't see a problem with my math as described in the prior comment. 1.16A, 122.4Vms, 0.63PF = 142VA; 85W per leg (L1, L2). Eventually the sun brightens and my PF works its way to 1.0 at about 900W from the array.

Can someone explain what is happening?

Hi ART005 - My Enphase IQ6+ base system also exhibits the problem you shared. Since installing it in October 2018, the Production meter shows 'Production not balanced between phases ....' message in Toolkit during non-production hours and during low production periods (as you described.) Wiring is correct and everything is by the book. Tried everything - no joy.

I have an active case with Enphase with an electrical engineer on he India team looking at the Envoy data to analyze and root-cause the issue. It raised 5 cases that were a dead end.
When this is resolved, I will post what information I get from Enphase.

The Toolkit screenshots below show three scenarios of imbalance between L1 and L2. The only time the imbalance goes away is during peak production at high wattage.

1) Almost-balanced production @ 2.80KW
2) apparent low production on L1 during cloudy period and 288W
3) 0 watt production before sunrise

Screen Shot 2019-01-04 at 10.24.59 PM.pngScreen Shot 2019-01-04 at 10.22.19 PM.pngScreen Shot 2019-01-04 at 10.26.09 PM.png


The next screenshot shows the cumulative effect of the negative reading on the lifetime measurements;

So although the system is producing balanced electric output by all the inverters, the Envoy is tallying the negative values and reducing the cumulative total: 693KWh on L1 vs 966KWh on L2
This is the total from October 12 - January 4 2019

Screen Shot 2019-01-04 at 11.25.06 PM.png


If you are interested in my investigation and analysis, read further:

I attributed the phase imbalance to the fact that my grid (NY Con Edison) is 120/208V i.e. the phase difference between L1 and L2 is 120 degrees instead of 180 degrees as in a 120/240V system.

1) My first thought was that the Grid Profile must be the cause of the errant readings, so after the first case Enphase, they told me to set the grid profile to:

SAM-60-120-1304 ver. 2

Highly configurable 127v VRef, 60hz profile intended as a source for custom profile derivation, including Advanced Grid Functions.

No effect on production imbalance although the grid profile seems to be more tolerant in the case an older or noisy grid.

2) The second analysis is more ambitious as I considered that the IQ6+ micros are sensing L1 and L2 voltage, phase, and frequency correctly and correctly matching them and outputting 120/208V at 120 degree phase offset and 60Hz. BUT, and this is a big BUT, since the Envoy iQ is designed to measure production using only one CT on L1, the Envoy IQ must be assuming that L2 production is identical yet mirror image of L1 i.e. 180 degree offset to L1, except that since the voltage measurement between L1 and L2 as measured by the Envoy must be 208v not 240v, the Envoy internal math to derive L2 must be wrong.

I opened a case and spoke with Enphase about the phase measurement and the math, but I was told that the angle difference is 120 not 180. Hmmm! what else could it be?

I have no way to prove this out as these measurements are in the firmware and electronics of the device and the Toolkit UI does not expose them.

3) So I started reading about how Enphase IQ6+ could deal with a 3-phase 120V Y system where there are three 120V lines and a neutral, and the difference between any of two lines is 208V. I discovered that there is an Envoy IQ Commercial (https://enphase.com/sites/default/fi...nvoy-EN-US.pdf)
which supports 208Y/120 VAC three-phase and 220Y/127 VAC three-phase. Examining the installation procedure, it seems they use a CT on each of lines: L1, L2, and L3 - this means the Envoy does not assume anything about each line but measures everything. I can't prove this, but it logically follows that the math must be correct inside this model. I checked with Enphase, and they said, Yes - that this is my problem, that's the root cause, and only the Envoy IQ Commercial could fix it because I would have a CT on L1 and a CT also on L2 and the problem is solved. So I said to them I would like to exchange my ordinary Envoy IQ with the commercial one - they said no can do. I have to buy one - it's approx. $700 USD for one. No sir. I'll keep digging.

*Note: I relied on the Envoy-S teardown article found here: https://thecomputerperson.wordpress.com/2018/08/18/what-is-inside-the-enphase-envoy-s-teardown/

4) Then I came across the ENVOY-S and how it deals with 3-phase - seems the they did not have a commercial and ordinary version - the SAME unit could deal with single and 3-phase BUT there a MODE setting: SIngle Phase and Multi-Phase. Hmmmmmmm. Looking at my Toolkit again, I saw that I have Single-Phase (L-L). Also, examining the Production and Consumption connections blocks on Envoy IQ, Envoy IQ Commercial, and Envoy-S, I saw that the terminals are the same but the wiring instructions are different. So I figured that if I installed a second production CT on L2 and wired it to he Envoy IQ like I would wire the Envoy IQ commercial, perhaps the MODE would automatically get set to Multi-Phase. So I bought an Enphase production CT off eBay and installed it - no go. The readings were wrong as the wire connections made illogical inputs to the computer and firmware. I tried every combination of wire arrangement - no go.

I was so convinced that Enphase would NOT manufacture two different products - I though they must be sneakily selling the same product as two different ones by adding two additional CT to the commercial version. This kept me chasing after a way to set the MODE to Multi-Phase. I could not find a way even after resorting to CSS and HTML page manipulation in Toolkit and API calls. I emailed Enphase but go no answer - did not open a case. I gave up on that path although I still have the second CT on L2 but not wired to anything.

5) Finally, I opened the currently active case with Enphase asking them how to set the phase difference to 120 degrees and sent them the images above. The screenshots did the trick and the engineer said he will investigate this because everything checks out - he had not noticed the imbalance.


Interestingly, another Enphase engineer solved a problem with Toolkit which showed persistent errors on 13 out of 27 microinverters that never reset during production hours. I had never gotten green status on all micros since first day they went online. That problem took 4 cases to finally reach the right person, and the right root cause. They released an Envoy software update for that one.

[ART005]

Enphase seems to have a bug where they sum the amperage values for each phase and then multiply that by 240 to get VA. That's wrong and results in a VA value for TOTAL CONSUMPTION that's upwards of twice what it should be.

My Enphase IQ6+ based system is a few days old. I'm 46 (edit not 43) degrees north latitude with a big hill to the south. About 1 month +/- the winter solstice the sun rises and sets a little more than 1 hr later/earlier than sun charts predict for my location. So I have a lot of indirect sun during that time. From about 8-11am my PF is 0.5-0.7 while a 2.5kWp array puts out ~300VA and ~180W. I had anticipated the 300VA, I'm sure it's correct. But I never realized I would get docked again by PF in the conversion to Grid power. I don't see a problem with my math as described in the prior comment. 1.16A, 122.4Vms, 0.63PF = 142VA; 85W per leg (L1, L2). Eventually the sun brightens and my PF works its way to 1.0 at about 900W from the array.

Can someone explain what is happening?

[sensij]

Long story short, I've got things wired correctly, but Enphase isn't calculating VA correctly in the Total Consumption category.

Thanks for the detailed update!

[rszimm]

So I thought I'd update people on what I found just in case someone else runs into this problem.

first, to sensij, the readings here under production are made with an actual CT off the panels. The readings under NET are made with a pair of CTs on the full house wiring (includes production and consumption). The readings under "TOTAL CONSUMPTION" are derived from the other two sets of readings.

Here's what I've found: the values for amperage, VA, wattage and PF on production are correct. The values for amperage, VA, wattage and PF on NET are correct. The values for wattage and amperage on "TOTAL CONSUMPTION" are correct. The values for VA and PF on "TOTAL CONSUMPTION" are not correct. Enphase seems to have a bug where they sum the amperage values for each phase and then multiply that by 240 to get VA. That's wrong and results in a VA value for TOTAL CONSUMPTION that's upwards of twice what it should be. That obviously throws the derived power factor way off, which is what drove me to look into this in the first place. It's surprising that such a bug still exists with a product that's as well established as this, but the VA isn't really used for anything and isn't actually displayed anywhere but in the installer toolkit, so maybe it shouldn't surprise me.

Long story short, I've got things wired correctly, but Enphase isn't calculating VA correctly in the Total Consumption category.

[sensij]

I don't really understand how the system is configured. The output suggests there are four data sources... production from the inverters, and then a type "eim" reporting production, a type "eim" reporting "total consumption", and a type "eim" reporting "net consumption".

On what circuits are the CT's actually located? On most systems, "total consumption' can only be derived from the production circuit and net consumption at the connection to the grid, not measured directly.

[bcroe]

You may have current leading or trailing the voltage for poor Power Factor. The PF of the
inverter feeding the grid should be close to a perfect 1, loads could cause out of phase currents.
If your voltage reference connection has a problem, a false PF reading could result.
Bruce Roe

[rszimm]

I think the negative power factor happens when you're sending power back to the grid, but I could be wrong. Anyway, I tried reversing both the production CTs and swapping their phases. Here's what I get now:

{"production":[{"type":"inverters","activeCount":28,"readingTime" :1515273265,"wNow":171,"whLifetime":9601464},{"typ e":"eim","activeCount":1,"measurementType":"produc tion","readingTime":1515273272,"wNow":2550.319,"wh Lifetime":9536358.584,"varhLeadLifetime":0.758,"va rhLagLifetime":2290273.033,"vahLifetime":10872994. 47,"rmsCurrent":21.212,"rmsVoltage":245.304,"react Pwr":498.904,"apprntPwr":2602.068,"pwrFactor":0.98 ,"whToday":28939.584,"whLastSevenDays":264992.584, "vahToday":32710.47,"varhLeadToday":0.758,"varhLag Today":6692.033}],"consumption":[{"type":"eim","activeCount":1,"measurementType":"t otal-consumption","readingTime":1515273272,"wNow":689.6 36,"whLifetime":15437803.168,"varhLeadLifetime":23 45897.04,"varhLagLifetime":4137920.844,"vahLifetim e":16759706.691,"rmsCurrent":4.634,"rmsVoltage":24 5.313,"reactPwr":-1236.266,"apprntPwr":1136.884,"pwrFactor":0.61,"wh Today":15756.168,"whLastSevenDays":232558.168,"vah Today":32086.691,"varhLeadToday":8962.04,"varhLagT oday":6828.844},{"type":"eim","activeCount":1,"mea surementType":"net-consumption","readingTime":1515273272,"wNow":-1860.683,"whLifetime":10563982.827,"varhLeadLifeti me":2345896.281,"varhLagLifetime":1847647.811,"vah Lifetime":16759706.691,"rmsCurrent":16.578,"rmsVol tage":245.322,"reactPwr":-737.361,"apprntPwr":2035.162,"pwrFactor":-0.91,"whToday":0,"whLastSevenDays":0,"vahToday":0, "varhLeadToday":0,"varhLagToday":0}],"storage":[{"type":"acb","activeCount":0,"readingTime":0,"wNo w":0,"whNow":0,"state":"idle"}]}