Conext ComBox aparently showing incorrect Load Watts

rayrassi · April 2015

Ever since I installed our system which is now about 7 months I have been intriged by the load watts registered by the system compared to the battery charge watts. I am new at this but something tells me it is impossible to consistently produce more than twice the kW than you are consuming unless they were being stored somewhere. I have a feeling I have a setting setup incorrectly somewhere or something wired incorrectly which could explain this. I am hoping someone has run into this or knows enough about this equipment and settings that can point out what I am missing. I have double checked the wiring and setting but cannot find the problem.

When I use the utility bills from before I went solar I figured a daily consumption of average 3 kWh.

We have 6 295w panels from DMEG DM295-P156-72, MPPT60-150 controller, Conext SW 4024 Inverter/Charger, 8 Crown 235AH 6v batteries.

I can post the settings page of any of the devices if you wish.

Attachment not found.

Photowhit · April 2015

Might it be that the load represents energy produced by the array that went to load rather than storage?

Energy out of the battery should be about 80% of the energy into, perhaps you're just seeing the number scewwed by the loads used when energy wasn't stored?

rayrassi · April 2015

Well, considering that this is a off grid system all the energy that is produced should in theory be consumed, I think. So why the load numbers show about half the energy produced?

mike95490 · April 2015

The com box does no measurements - it only reports, the inverters and charge controllers do all the measuring. In my setup, I manually log the charge controllers, and the inverter readings.
With my NiFe batteries, there is not a lot of efficency, which is why I had to add a 2nd array. This little sample shows my charging sources, loads and fuel consumption,
Attachment not found.

Photowhit · April 2015

As Mike said, batteries are not 100% efficient, only about 80-85% so you put 100% in and take 80% out....

The numbers will be better for energy consumed while being produced, but inverters are not 100% efficient either, more likely 90-96 with an Outback, so if they are measuring output in AC it will take more input to create...

So the energy is being consumed but perhaps lost to efficiency, Indeed off grid is all about waste! You will likely loose more in potential power than you will store on most days. Also the reason off grid people like to setup their systems so loads run when the batteries are topped off.

DMJ72 · April 2015

Photowhit wrote: »

As Mike said, batteries are not 100% efficient, only about 80-85% so you put 100% in and take 80% out....

The numbers will be better for energy consumed while being produced, but inverters are not 100% efficient either, more likely 90-96 with an Outback, so if they are measuring output in AC it will take more input to create...

So the energy is being consumed but perhaps lost to efficiency, Indeed off grid is all about waste! You will likely loose more in potential power than you will store on most days. Also the reason off grid people like to setup their systems so loads run when the batteries are topped off.

I agree 100%, but the big difference in the numbers makes me wonder if the CSW/MPPT/Combox has a bug in the accounting of the power generation / utilisation. Here is my off-grid system for comparison :

Attachment not found.

rayrassi · May 2015

Thank for all the replies. Looking at DMJ72 I guess I'm not the only one who shows the disparity. Like he said I understand there is some loss due to ineficiency but more than 50%??? Does anyone have contacts at Xantrex that could shed light on this?

Dave Angelini · May 2015

rayrassi wrote: »

Thank for all the replies. Looking at DMJ72 I guess I'm not the only one who shows the disparity. Like he said I understand there is some loss due to ineficiency but more than 50%??? Does anyone have contacts at Xantrex that could shed light on this?

You either have programmed it wrong, have defective equipment, or alot of loss from installation. One way to test it is to use the SCP harvest data from your solar, or read the data from the mppt-60. To do this, you must have charged the previous day. You should look at the harvest in KWH at the same time each day for a couple days. The number you get is all the energy to charge your battery, the power you used overnight, and the power you used that morning up until you went to float. Compare your data with the combox

Ganglefang · February 2016

Using the energy balance equations, (energy in equals energy out) I have noticed when I use the numbers from the main Combox display that you have to multiply the battery energy (Watts) by about 10.3 to make the equations balance. Just saying. I know, E=M*C**2 and W=E=V*A. 12v, 240 v, etc. I haven't figured it out yet either.

Dave Angelini · February 2016

Energy in + loss = energy out.

Ganglefang · February 2016

Too true. But if the discrepancy is losses, then my system is sometimes losing several thousand watts. Then at other times it is creating as much. Hmmm.

Dave Angelini · February 2016

You really have to make an effort to describe your system if you want decent advice.

zoneblue · February 2016

Well lets look at what might be a reasonable set of numbers.

Assuming we are talking about CC output over AC load consumption...

- Inverter effciency, 85% typical, rule of thumb
- FLA charge cycle effciency (80%) again rule of thumb. (depends on how deep your cylces are, and AGM/FLA).
- BOS losses, should be pretty minimal if post CC cables are short and fat, lets say 3% just for a place to start.

So assuming you use half your power outside sun hours, then:

Overall effciency= 0.85 * ((1 * 0.5) + (0.8 * 0.5)) * 0.97 = 74%

In practical terms here, while i only track loads at the DC stage, i get:

select
    count(*) as ndays,
    sum(whtotal),
    sum(whload),
    sum(whload)/sum(whtotal)
from classicdaylogs

ndays  sum(whtotal)  sum(whload)   sum(whload)/sum(whtotal)
-----------------------------------------------------------
605    1494620       1401104.9     0.937432190121904

94% of what the classic produces gets consumed by loads (pre inverter). That just tells you that a) the charge efficiency of AGMs is quite a bit higher than FLA, and b) we try to not run so much stuff after dark, and c) we turn the VFX off at night (which increases overall conversion effciency). As for the inverter Id need to run some tests to say what my typical VFX conversion loss is.

Thats probably more than you wanted to know. tldr: theres something screwy in your setup somewhere.

Ganglefang · February 2016

Right now, my system is selling 4305 watts. Input is 239 (SOLAR) and 550 (Battery.) No generator. 550+239 = 789. So 4305-789 = 3516 is coming from somewhere. I may patent this system. Fact is, if I multiply the battery watts by about 10 or 11 and ignoring losses, the balance works out. My point is that there is something wrong with the battery watts numbers.

On grid
Conext XW+ 6848 Inverter
2x Xantrex XW 80 600 MPPT Battery Charge Regulators
Conext COMBOX connected to the local network
Ver03.00BN0704
Conext Battery Monitor
Xantrex AGS Automatic Generator Start Control
Generac 20KVA generator
20x Batteries 8A4DLTP-DEKA 210Ah batteries
24x Roof Mounted REC 255W panels

Dave Angelini · February 2016

The first COMBOX system performance screen, the one that it boots with shows arrows on the direction of power flow.
You should see power flowing to the grid if selling, you should see power flowing to your loads, should see solar flowing in if the sun is out, and you should see power flowing in or out of your battery.
You are not seeing this????

Ganglefang · February 2016

Yes, I see this. I watch it all the time. It is this display that I am talking about. This is where I got the numbers stated in my previous post. I take a screen shot and add up all the numbers. And what I am saying is that the Inflow doesn't equal the Outflow, under various situations of Selling, Buying, Charge and Discharge. The energy flowing in/out of the battery is off by a factor of 10-11.

I notice my system is no longer new, so I need to update the firmware(s). But looking over the release notes, I don't see this question is addressed. Nevertheless, I will update all devices. And send a query to Schneider. I will not hold my breath on either of actions. It is also possible they sent me the wrong shunt, but I inspected it when we installed it, so that also is a long shot. If they would send me the code for the firmware I could debug it for them. That is really a long shot. For now, I just multiply the battery energy in-out by ten. Lord knows we have to do a lot of other workarounds with this software.

--------

On grid
Conext XW+ 6848 Inverter
2x Xantrex XW 80 600 MPPT Battery Charge Regulators
Conext COMBOX connected to the local network

· Application Version Ver03.00BN0704

· Application Build Date 2015-01-29_10-42-42

· Bootloader Version Ver01.03BN0118

· Bootloader Build Date 2014-12-16_12-18-3

Conext Battery Monitor
Xantrex AGS Automatic Generator Start Control
Generac 20KVA generator
20x Batteries 8A4DLTP-DEKA 210Ah batteries
24x Roof Mounted REC 255W panels

Dave Angelini · February 2016

Hmm, probably settings, it usually is! But, if the shunt is wrong all bets are canceled. Take it out! You do not need it with an XW !!!
There is always inversion loss and it will vary at different battery voltages and loads, (especially low loads) I often see 100 to 300 watts of this on my 5KW system.

It is not normal and there are not issues with this, I have the same firmware but no grid or generator.

zoneblue · February 2016

Any monitoring is a proxy for something real. You are right to question the shunt, as you ought to question all the data. Checking the data from time to time against a known good multimeter or other source will let you verify the monitoring metrics, before relying on them. For example we get people here all the time saying my bank is at 95.2% because my battery monitor says it is. Might be, but more than likely is at least a little off, and might be way off.

Dave Angelini · February 2016

Yep Zoneblue! The XW voltmeter is plenty accurate for this. In his example if he looked at the direction of power at the battery (in/out)
or the voltage (is it being discharged) something is supplying 4,305 watts to the grid with only 239 watts of solar.
If that was a sunny day, why is the solar so low? If there was a system control panel, is the data the same as the combox?
Maybe do a complete DC power down for a few minutes and restart? Grid people often have power glitches do weird things.

You and I do not have those problems !

Ganglefang · February 2016

It was an overcast day, and i have checked currents and voltages independently using voltmeter and ammeter. I don't think I'll throw out the shunt.

Still puzzled. I will report back after I have installed updated firmware and talked to Schneider.

Ganglefang · February 2016

Image: https://us.v-cdn.net/6024911/uploads/editor/ut/d5dri34z46zk.jpg

Ganglefang · March 2016

Apologies for not following up on this. Shortly after my last post, the Battery Monitor went dead. Once I got that replaced, my ComBox went dead. I am still resolving these problems. Schneider has been good about replacing parts, but each replacement cycle takes 2-3 weeks. (The energy equations still don't even come close to balancing.) As to throwing out the shunt, then I might as well throw out the BattMon, and while I am at it the fuel gauge on my car.

Dave Angelini · March 2016

You do not have to throw out the shunt but rather lose it for sometime for troubleshooting. My guess is you are powering loads that are in a sub panel or something is heating up such as wiring or components. Since you are on the grid you could easily have damaged equipment from surges. You have SPD's right?

Ganglefang · March 2016

I have a problem believing all the lost and gained power are due to internal losses. Einstein would be turning in his grave. IMHO There is something wrong with how they compute power. I have used my thermal camera to check for hot spots, but found nothing.

We have lots of SPDs, including whole house and one on each major component including one each on the chargers, one on the inverter, and one on the PV breaker panel. And we have lots of computers and other electronic equipment, and a recording monitor on the grid and PV power. No spikes, and the only equipment failurees have been to a couple of phones. Even the Combox is on an additional UPS/SPD. We are blessed with pretty clean power here from LPEA.

In any event I am struggling with getting the Combox replaced. Since it was installed by a certified installer, I do not have the invoice, and seems to be a problem.

Ganglefang · March 2016

I meant to add this to my last post. In the system shown above, we have 7414 watts going into the system and only 2474 coming out. That would be almost 5000 watts missing due to internal losses? Other times I see the opposite with the system "creating" thousands of watts. I have a spreadsheet documenting these losses/gains. I guess nobody else is COPD enough to notice.

Dave Angelini · March 2016

It is obvious that something is wrong but keep in mind the 45,000 (my guess) systems out there that do not have your problem.
Get some meters out or someone who knows how to use them and troubleshoot. Your battery voltage in your plot is not what you should see at 99% state of charge.
This will not be hard to find and if he were around, Einstein would tell you that.

Ganglefang · March 2016

I finally got replacements for my Combox and BattMon, and I will be installing them this weekend. But I am curious, has anyone else noticed similar anomalies in the data reported by their Comboxs?

Ganglefang · August 2016

It's now mid-August. I have been through 3 comboxes and 2 battery monitors. I added a system control panel since the other boxes seem to lack reliability. (Yes, I have surge protectors on everything: MPPTs, inverter, both breaker boxes, line input, generator, in some cases several in series.) The Combox continues to report data inconsistent with my inverter and my meters. I pointed this out to Schneider in two emails but never got any response. (Yes, I checked my junk mail box just to be sure.) I guess I am the only one noticing these problems, probably related to the fact that most Microsoft Updates contain errors that nobody else notices(?) OCD? At this point I am closing the issues, and returning to my day job as a graduate EE (50 years) designing nuclear weapons.

Dave Angelini · August 2016

Give me your case number and I have a friend there who will look into it. If you do not have a case number how can you expect anything to be resolved. What does your installer say?
What do you think is happening as an EE?
5kw in from solar in your plot. and 2kw in from the grid and a 2kw load. What does your utility bill say?

WaterWheel · August 2016

Ganglefang,      I also run a 6848 with an 80-600 controller, no combox but an SCP and Conext battery monitor.     I don't sell back to the grid but similar equipment.     On post #21 if you reverse the power coming in from the grid the numbers come much closer to balancing.

I do agree that the Conext battery monitor is less than accurate when measuring SOC.      While I wouldn't expect it to be dead accurate mine accurately matched discharge levels (as measured by evening and morning SG measurements) but during charging the next day it shows 100% SOC when the batteries are about 88% full (as measured by SG).     This error compounds some after a week or so until after 2 weeks the SPC can show 100%. but SG was 2.204 (this occurred before dialing in a new controller).      I have tested this several times equalizing to 1.277 SG and the following day with the batteries charging and a SG around 1.255 the meter will show 100% SOC.

Tom Ribaudo with Schneider customer service has given me exceliant help.      Give Dave Angelini your case number and see if he can get you the help you need.

new2PV · August 2016

If the battery monitor has a weak connection, it will make a mess of the battery watts readings, maybe this needs to be looked at?

Conext ComBox aparently showing incorrect Load Watts

Comments

Categories