Four XW6848s, 17 Discover AES 6650, Conext Gateway, eGuage, Load Calculation Differences

Gullo

Good morning Forum, I have posted about this before, and you were super helpful. Today I'm going to try and include pictures and more complete detail. I am experiencing a huge variation between what Schneider reports as my loads and what was calculated as my loads prior to my install AND what is continually reported by my permanent eGuage as my loads. Schneider regularly reports that my daily total loads are in excess of 35kWh, and my eGuage reports 20kWh. Please see the excel spread for one day. I keep these records EVERY day.

I'm sorry this is blurry, but it's a laptop screen, but if you zoom, you should see detail. I keep 3 primary calculations. 1. The hourly reporting from Gateway, 2. The hourly reporting from eGuage, and a third that seems to isolate the problem, and that I will try to illustrate in a photo below - Hourly reporting from AC Loads from all 4 breakers on the inverter side (basically, the loads that the inverter is supplying)

The Excel shows, Loads, eGuage Loads, and Inverter Load CT totals.... Please notice that Loads (which is a Schneider and Discover product represented in Gateway) is MUCH higher than the eGuage (which is the result from the eGuage right at the load center). What is interesting is that Inverter Load CT totals agree more closely with Gateway/Discover.  Which is to say that somewhere between the AC Load breakers in the PDP and the actual load center, there is significant loss/resistance/heat. From this point, there are 3 downstream "devices" 1. The AC Combiner box, 2. another combiner box that allowed the installer to configure the system to operate in AC pass-through and to utilize the Automatic Transfer Switch, and 3. the Automatic Transfer Switch. I tried to post the three devices below, but it pasted above. I have taken FLIR photos, and I see no resistance in the primary combiner box, nor any of the other system devices. I would love to find out why my loads are so different. It's a huge problem! As it stands, I am depleting my batteries nearly twice as fast as I expected, which suggests that 113050 kWh worth of usable energy is not enough - which is a ludicrous concept!

mcgivor

One possible cause for the discrepancies in reading may be due to CT crosstalk, having the CT's packed close together, and or parallel conductors in close proximity to the body of a neighboring CT, can influence the values of the output. See link

https://ctlsys.com/support/current_transformer_crosstalk/

mike95490

Do any of the measuring devices measure VA or Watts ? Any inductive loads can "seem" like more power being consumed

Gullo

McGivor, I'll slide the CTs down the inverter load wires to get them a bit further away from one another, and I'll let you know                     Mike, I know that Combox was able to be changed to VA for measuring, but I'm not sure Gateway can - and I definitely don't know if I can measure volt amps with the current transformers from the eGuage. I'll call them Monday and find out. I don't have many heavy inductive loads. I have a couple of super high efficiency ceiling fans that run at 20 Watts, and I have a geothermal heat pump (Hydrotemp V-Star 5-ton) that runs (this time of year) between 400 and 800 Watts when it's running. It's a regular house - LED lights and TVs etc. I have a woodworking workshop, and that would be very inductive, but I haven't been working out there since I got sick.  

Gullo

Here is a visual representation of the Wattages between 12p and 1p yesterday... The Inverter side of the 60Amp AC breakers has a total load of 1590 Watts. Those "travel" through the breakers over #6 THHN to 4 60AMP breakers in the AC combiner box above, and then to the Grid combiner box, which splits the AC so that it can be sensed by the Automatic Transfer Switch 12v relay and open or close the 200AMP contactor, and then to the load center, which is reporting 1070 Watts for that hour ---- These wattages are also being reported by the Schneider equipment and reported in the Conext Gateway as 1900 Watts..... SO, what's happening is I calculated PRE-installation Daily loads at roughly 20kWh per day in months with very minimal HVAC. After installation, I have been reporting daily loads averaging more than 35kWh. As you know, battery pack sizing is based on consumption. I based mine for 5 days of autonomy, but I have been lucky to achieve 3 days. SOOOOOOOO I bought and installed an eGuage, and the eGuage agrees with my original load calculations, and is in total disagreement with the Schneider equipment. It's an interesting problem, and I'm sure the answer is discoverable. Naturally I assumed resistance was the problem, so I got a FLIR camera and as you can see there is no heat in the most likely candidate, the grid combiner box. I saw no unusual heat. Then I took a pic of the entire system, and still there is no unusual heat. You can see that the inverters are the hottest part of the system, and they're perfectly normal. I also looked at the inverters with the covers off, but I didn't take a picture of that.  

Gullo

One last detail..... This is an in the moment snapshot of the individual eGuage registers. What you see is 3 different "Totals". These are points in the system where I have current transformers... Load Side AC Combiner (Yellow) is the combiner box above the inverters. INV *AC Load (Orange) is where I have a CT on all 8 of the inverter side of AC Loads before the electricity is sent to Yellow. Red is the eGuage grid inputs. It is considered by eGuage to be the most accurate read of total loads. At the moment that I snapped this picture of the registers, Schneider was reporting a load of 1500 Watts.  So we have 4 different reads of the same load...  Schneider totals are 1389.90 (by the eGuage) and 1500 (by the Gateway) and 833.6 versus 875.3 totals of the L1 and L2 at the combiner box and at the Load Center. 833, 875, 1389, 1500 That's a big spread. Any thoughts? Anyone want me to provide a different view? I would love to tinker with this and figure it out.  One thing I do notice is that the two LEGS are not balanced. Apparently I have a heavier load on Leg 1 than on Leg 2 - for what it's worth.  But I'm pretty sure that could not account for an 80% difference between kWh readings.

Estragon

So if I'm understanding this correctly, the big discrepency is between watts as measured on the 8 individual L1/L2 source wires (measured internally in inverters, or externally at inverter output breakers) vs on combined L1/L2 load side wires.

Maybe a silly question, but do the CTs measure actual current (irrespective of voltage phase angle)? The wattage is after PF losses etc? Is PF measured somehow, or calculated using assumptions (eg sine wave shape, harmonics)?

I don't know enough about this, but I'm wondering if the combining is somehow distorting the measurement or wattage reporting.

That your batteries are apparently at states of charge reflective of the higher wattage numbers, and the lack of apparent wire/connwction heating, seems to me to suggest the issue is somewhere in PF.

Have you tried comparing the discrepency with and without various loads (eg heat pump) connected?

Gullo

Yes, I have minimized loads and tried "known" loads, and the discrepancies still exist. I am wondering, although the book does not suggest it, are the inverters using energy to convert energy? I mean how many watts are the inverters using just by being on and inverting from battery DC to AC?  It seems unreasonable that this conversion would be as much as double the reported load center requirements, or my NABCEPs would have told me about that enormous discrepancy when I was planning my battery size.  I mean, is this normal? Does anyone else measure their AC loads compared to what they're "consuming" from battery?  

Estragon

My guess is an inverter your size would use on the order of 50ish watts just being on with little no load. Peak AC/DC conversion efficiency would be on the order of 85%, but would be less at very low load.

Constantly running, small loads can definitely be a problem running on a relatively big inverter, but that doesn't explain the discrepency in measured AC wattage.

Gullo

Would one way to test that be.... Turn off 3 inverters and let the system run on just one?  I'm going to try that.