Archived Forum

I've just found your later posts on the other thread (somehow, I missed these a month ago).

First, a misconception to be dispelled. There is no inherent current limit for the emonTx - it depends solely on the CT you use to drive it and the burden. So if you have a 1000: 50 mA CT and the in-built burden, it will read 1000 A with the appropriate scaling in the software.

Now to this problem. You need to find out what your meter LED flash represents. There are principally two ways that I can think of, experiment or paper-chase. The experimental way is to wait until your plant is shut down overnight or for a holiday, then put a known load on and count and time the flashes. The paper-chase is to find out what the meter characteristics are, what those meter displays actually mean, what the ratio of those big CTs are, and work through the sums. The energy bill might have some information that helps. Even better if you can talk to a friendly engineer at your electricity company. I've no idea whether you will be able to find out everything that's necessary, so the experimental way might be the only practical way.

Certainly, the top picture says it's kWh, but they might have forgotten to scale it for the CTs, so it's 8 x (the CT ratio), which, if the CT is 200 : 5 say, makes much more sense of your 8 kWh ( x 40 = 320 kWh). That's just a theory, so treat it with caution.

The 0.48 / 0.59 kW display:  If it doesn't change when the load is cycling, could it be maximum demand - again scaled by the CT ratio?  Some suppliers charge a penalty if the maximum demand goes above an agreed limit within a specified time frame, usually half an hour. You might find this resets periodically.

The 0.919 could indeed be power factor. This is the ratio of real power - the power that a resistor (heater) dissipates, compared to the power that you calculate when you measure voltage and current and multiply the two numbers. For a resistor, the value is always 1 (by definition!). For a reactive load, like your compressor motors, it will be less than 1, and 0.92 is good. For a very small induction motor, it can be down to 0.6 or so. A poor power factor is bad because it means you have to have fatter cables to deliver a higher current for the same amount of power. Some suppliers charge extra if you have a poor power factor.

It took me about 5 s with Google to locate a manual for your meter:
site.ge-energy.com%2Fprod_serv%2Fproducts%2Fmetering%2Fen%2Fdownloads%2Fkv2ci_oi.pdf
I think you need to study that. I haven't done so.

That should tell you what your meter is displaying. I mentioned in the other thread about putting your own CT on the meter CT secondary side. Is that possible, or can you get a clamp ammeter on there? If you could, that might help to resolve what the meter scaling is. Let's assume those big CTs are 200:5  Then if the current into the meter correlates with the meter kWh incrementing, and the LED flashes (which I guess might actually be more like 1 Wh but again scaled by the CT ratio) then all you need to know is the metering CT ratio.

Having quickly looked at the manual, you also need to know k_t. It should be underneath the paper label on the front of your meter - see Fig. 1-1 of the manual. It would be useful to know anything else under there too.

The manual says the meter defaults to 1 pulse = 1 Wh, as I guessed. But that doesn't necessarily mean yours is set to the default.

Also available is a comprehensive data extraction kit: http://wattmetrics.com/products/supportedpowermeters/GEkV2c.aspx is the first I noticed.

Ah, thanks for your input, Robert! (and sorry for the delay - honestly I'm not sure if I set up the email notification on this forum correctly.)

Okay, I have some new information to share with everyone based on Robert Wall's feedback.

So experimental vs paper-chase. This is a great way to organize the information-seeking process. I think there may a few things holding me back from being successful with either tactic. I could be wrong and these issues might be inconsequential. 

Experimental

• Even at night, I can't totally cut off every load in the building. Alarms, monitors, aquarium, etc. Even during the night hours, our building is "on".
• I might not have a clamp meter large enough to go around the main cables. If I did, I still might have an issue getting inside the "sealed" exterior box that houses the monster CTs. (These boxes have tamper-evident clips on the doors, but I was able to obtain a a few photos by prying it open ever so slightly.
• I'm not opposed to calling up the utility company and attempting to track down an engineer (or anyone beyond tier 1 support staff). I will look into this - great suggestion.

Paper-chase

• Gathering data about the hardware in use in order to figure out what they might be telling us.
• This is the (quite spooky) inside area of the box that contains the 3 CTs.
  • 
• I was able to get a close-up of one of the metal plates on the CT. This might be a critical piece of the puzzle, allowing us to no longer make certain assumptions.
  • 
• Another required piece of information from your look at the manual. The value of k_t on the meter sticker is 1.8. (It was hard to make out in the photo I snapped, so I typed over it.)
  • 
• For good measure, here is a representation of the electric supply found on the building's exterior
  • 
• Also, a copy of the interior layout (from a previous post)
  • 

I had taken a look at the meter manual as well and was discouraged when I read that they could all be programmed differently. I believe this is the reason they do not have an index of meter screen numbers (05, 07, 18 from the original post) coupled with their relevant descriptions.

I have looked into a similar data extraction kit, but there doesn't seem to be any that I can purchase outright. Unfortunately, the WattMetrics system has discontinued sales.

Does any of the provided information reduce the number of variables we have to work with, or otherwise shed some light on the situation?

Thank you for your insight!

That does indeed help.

There's absolutely no reason why you can't put your own 5 A c.t. on the secondary wire out of the meter c.t.  But it must of course be a split-core one! You must never open-circuit a c.t. under load. If your c.t. is a 5 A : 50 mA (say, to give you a standard emonTx input and burden resistor value), then you have effectively got a 400 A : 50 mA one - you get 50 mA out of yours for 400 A through theirs. Knowing their CT ratio has removed all doubt about the calibration at that end.

"The value of kt on the meter sticker is 1.8. (It was hard to make out in the photo I snapped, so I typed over it.)" It was impossible in the original photo because there was another sticker over the top of it! Now that we know Kt also helps, it says you get 1 flash every 1.8 Wh. Therefore 4025 flashes = 7.245 kWh on the meter (that correlates with your increment of 8)  x 80 (for the CT ratio) = 579.6 kWh consumed. That's way above your daily average so that apparently is not correct, but you have more information that should allow you to check - a good source would be old bills and meter readings, assuming they have been recorded.

[Starting a new thread after your first post here http://openenergymonitor.org/emon/node/4895  is a pain because it means I've got to go and find it each time - please don't open yet another thread on this topic!]

And as far as I know, email notifications have been broken for some time.