Hi, I have a design question and would appreciate some guidance.
I’m looking at using IoTaWatt devices to monitor power consumption across 31 x 3 phase circuits. The application is for a 10 storey 50 year old apartment building in Australia. Each level has a 3 phase circuit that feeds all units on that level, plus we have a single 3 phase circuit for common building power.
So all up I will need 3 devices and 33 CTs.
We only want to monitor the 3 phase feed and not individual apartments. It’s a 50 year old building, so we want to monitor and make sure the phases are balanced, and to proactively plan and manage upgrades to each level. We are also looking at this from a safety point of view so that individual phases are not overloaded where they potentially overheat.
Now, I currently have 1 device setup as a poof of concept and have implemented for 1 level (3 x 100 AMP CTs) plus the common power (3 x 50 AMP CTs). This is working well and I can demonstrate it’s value for the Owners Corporation from a building management point of view.
So now I am about to purchase another 2 devices and all required CTs to complete the project.
From a configuration point of view, would it make sense to assign each device to a phase? In other words I split the 3 phases across the 3 devices. So 1 device would be Red circuits, 1 for the White circuits, and the other for the Blue circuits? This would be handy as then I only need a single VT for each device.
You don’t say whether you are using derived reference or direct reference in your present setup. The difference is that derived does not take into account variations in voltage between phases, so Watts may be different to the extent there is a variation. Generally speaking, if one of the two “derived” phases has a voltage that is 2% greater than the reference phase, the power (Watts) will be 2% greater.
However, you describe an effort to monitor these circuits so as to balance and avoid overloading. That could easily be done with slight variation in Watts, moreover, the best metric for that is Amps, and that metric will be unaffected by voltage variations.
If all of these IoTaWatt units will be in the same place, then you can use a splitter for the AC reference(s). You can use one reference split to the three units for derived reference, or three VTs split to the three for direct reference. The only caution with using splitters is that each IoTaWatt must have it’s own independent USB power supply, using a USB hub or multi-point charger will not work properly.
As to whether to use one IoTaWatt for each leg of the three-phase, that’s a matter of your preference. If you want to be able to comparatively look at all of the loads on a particular phase with Graph+, that is a good strategy. If you want to be able to aggregate the combined usage of all three-phases by floor, then measuring all three in one IoTaWatt is a better approach.
Ideally, you would upload the data to influx or Emoncms to aggregate all of the data and it wouldn’t matter which approach you adopt.
Hey, thanks for the response.
At present I am using a derived reference for the proof of concept, but want to move to a direct reference to ensure total accuracy with whatever measurement is used as I want to be sure I design this right from the start.
Agreed that Amps is the main measurement I’m after for what I’ve described.
Yes, all devices will be in the main electrical room on the Ground Floor, which takes the feed from the main road. From there the phases are run to each level.
Using a splitter for the VTs is handy to know, which will reduce my costs and limit the number of GPOs I need to have installed.
I’m currently pushing the data to influxDB and then Grafana to get a nice dashboard for the Building Manager and Owners. And I can then build on that.
Thanks heaps for your response. It gave me a level of confirmation for my design, and allows me to implement either option as a preference.
Surely you would need bigger than 100amp CTs for the main incoming 3 phase ?
Thats only 72KW in total for the whole building ? 4 units per floor = 40 units = < 2Kw per unit unless my maths is very off ?
No. The main 3 phase power feed into the building connects to 3 x copper busbars, and each of the 11 x 3 phase feeds connect to the busbars. The 11 feeds are for each level plus the common building power. So I am connecting a CT to each of these. Hopefully you can sort of see this from the attached photo. Remember that this is a 50 year old building, so it isn’t pretty. So our goal is to understand when we need to upgrade each orange circular cable and/or balance the power across the phases on each level. We can do all this by locating the devices and connecting the CTs in this main Electrical Room, rather than physically on each level.
You will have 3 unused inputs. I do have 400A and 600A x 36mm CTs if you want to monitor the mains coming into those buss-bars.
AAh Got it - my bad.
If you are going to push all of this data to Grafana/Influx - then there is a good write up here by another Aussie
Follow through - he links to another post also where he talks about tags etc
Yeah, I’ve been following Brett’s articles. Very helpful as well as the tagging. I’m using a Raspberry Pi 4 with the iotstack.
Also your reference on another thread regarding the use of a GL.iNET GL-MT300N-V2. The main electrical room is behind solid concrete with heavy doors. Existing WIFI signal is not strong and reliable, so I’ve run a CAT6 cable in there.
No worries - glad to help another Aussie !