Help me decide if IoTaWatt will work for our unique property

Hi There!
I just discovered IoTaWatt and have been reading as much as I can. I’m trying to learn if this system can do what I need it to.
I’m the caretaker for a very large private property. 3 main 3-phase electrical services, 72,000 sq ft of commercial style buildings. The owner is trying to get on top of the energy usage and see what changes we can make to reduce consumption. Sub metering is the only real way to get the data we need. I’ve been researching every metering product out there to tackle this project in a cost effective way.

Packet Power gave us a detailed estimate to meter our mains and all the mechanical equipment circuits. That came in over $26,000 with yearly licenses on top of that. That was a no go from the owner.
I made the recommendation that we start with a few Eyedro meters to monitor our building mains and the couple sub panels that contain all the mechanical equipment as that is the primary energy demands. Branch metering would provide the most detail but there is a cost/benefit trade off.

IoTaWatt looks very promising for our usage but there are a couple of things I can’t of yet find answers to:

  • Metering circuits with parallel feeders. The mains for the primary building service (2000 amp) are large. 4 parallel feeders for each phase. Eyedro told me you can put a CT (600 amp) on just one of the parallel conductors and in the setup select to multiply a CT by 4. Will that work, is that an option, with IoTaWatt?

  • Related to above, can you use rogowski coil CTs? Some of the sub panels in the buildings are again fed by parallel conductors. This makes installing split core CTs not an ideal option as some panels are very tight. (could make it work if IoTaWatt has the option to multiply the current from 1 feeder)

Photos speak volumes so if you have time and find it interesting check this album out and let me know if you think I can use IoTaWatt to accomplish our goals. This is all commercial level stuff.

You can do that with IoTaWatt, but there are better methods.

IoTaWatt doesn’t support rogowski coils, but looking at your very neat and organized feeds, the multiple parallel cables are nicely bundled and could easily accommodate large CTs. The largest that I stock are 600Ax36mm, which might work OK on the dual feeder circuits, but you would need larger for the three and four cable units. They are available on special order. One customer recently installed 1200A x 80mm. That line is available up to 1600A x 80mm. Here is a 800A x 55mm:

IoTaWatt can do three-phase by two methods:

  • Direct Reference - where three voltage reference transformers are used, one on each phase, to get exact voltage reference.

  • Derived Reference - where one voltage reference is used and that voltage is used for power calculations on all three phases.

Derived is much simpler and typically gives excellent results.

Each IoTaWatt can monitor up to 14 CTs, or 12 if Direct Reference is selected, so assuming you use big CTs on the mains, you could monitor 4+ sets of mains per IoTaWatt. When you get down to branch circuits, there are 50A, 100A, 200A, 400A and 600A CTs available.

From a measurement perspective, this is all very doable with IoTaWatt and some special order large CTs. Your biggest challenge would be aggregating the data into a centralized database to analyze it. Each IoTaWatt will store years of detailed data, but can only display and plot the circuits that it is monitoring.

Automatic real-time upload of the data to one or more database systems is supported. Both are open-source systems: influxDB and emoncms. Both also offer the ability to run your own server or to use an enterprise cloud system. Cost is low for the cloud systems, or zero for your own server.

It does however require quite a bit of setup to make it all work together.

This is a big project. One approach would be to start out monitoring all of the mains, and then use that data to decide where you want to drill down for more detail. is providing an online service that is designed for handling the data from multiple IotaWatts. This may help you organize in one place all the data you will be generating!


Thank you for the responses and insights. My boss and I agree with you on starting with a small system and expanding it as I gain experience.

Would you say that this is the better method to measure the large mains? Paralleling Current Transformers - Continental Control Systems, LLC
I would get a total of 12 CTs to monitor the 4 conductor 3-phase Main, correct?

Thank you for this information. Below are a couple examples of feeders I would like to monitor and the conductor size measured with a caliber.
“TAB Main” 3 Conductor bundle =65mm 1000 amp.
“Mech Panel” 2 Conductor bundle =45mm 600 amp
Do you stock CTs that would work for those?

It’s a better method, but can’t be done as described with IoTaWatt. To do that with IoTaWatt, The values of the internal burden resistors would need to be changed or external resistors added.

Yes, with that method, but I would need to change out the three burden resistors in the intended IoTaWatt inputs from 10 Ohms to 2.5 Ohms. Not a big deal but needs to be planned out which inputs are used for what.

I do not stock them. 1200A x 80mm would work for the 1000A 3 conductor. Back in April they were $150, lead time is about 2-3 weeks. Compare that to 3 400A CTs at $48 and it’s a wash.

The 800A x 55mm would work for the 600A x 2 conductor. Haven’t priced recently but expect about $110, same lead time.

I expect a 1600A x 80mm might fit your bundle of 4. I know that you said it was 2000A, but in general a service should be sized to operate at no more than 80% of the mains capacity, so 1600A should work. Price for the 1600A x 80 should be comparable to the 1000A x 80.

I can get a firm quote from my supplier if you tell me the number and type that you woud require. A big part of the cost is shipping from China, so as a special order the quote is dependent on quantity.

Thanks overeasy!

Is there a “best” method you would recommend for our use case in monitoring the mains?

I think I could get the large 80mm CT around the bundle as each conductor is 31mm. However I would like to make sure we do it with the appropriately rated CTs. We have features of the building that actually can come close to using the full capacity of the service if you ran them at the same time. It is not likely to happen but I would want to be safe and get the correct equipment. If a 2000+ amp CT isn’t available for the mains, then maybe a good compromise would be using the 1000a x 80mm CTs on a pair of conductors and parallel the 2 CTs per phase. Would that work better?

The larger CTs sound like a good solution for the 600A and 1000a feeders. Simplifying the setup and wiring, especially avoiding modifying inputs on the board sounds good to me.

We are very serious in wanting to move forward with a couple IoTaWatt’s to get started.
My list would be as follows:
IoTaWatt #1 - Derived Reference
2000A mains = Your best recommendation. (3 large 2000a, 6 1000a paralleled, or 12 600a, other?)
600A feeders = 3 800A x 55mm CTs
400a feeders = 3 400A x 36mm CTs
200A feeders = 3 200A x 25mm CTs

IoTaWatt #2 - Derived Reference
1000A mains = 3 1200A x 80mm CTs
400A 2feeders = 3 800A x 55mm CTs
200A feeders = 3 200A x 25mm CTs

I will get a quote on the large CTs and PM that to you.

Question about multiple mains options.
I’m not an expert by any means, but like you, have been doing a lot of reading.
Here are what I see seem like the options.
Adding comments for clarity is welcome.

with the 4 A-phase lines, 4 B-Phase, 4 C-Phase

  1. one large CT for all 4 lines on A-phase, 1 for B and 1 for C, all power measured. (preferred method)

#2) one CT on one line, assume reading x4, 3 go unmeasured. (adds uncertainty)

#3) one CT on each line, each CT fed back to one channel, 12 channels.
(complicates data logging)

#4) one CT on each line, modify units so that the 4 CT’s on A-Phase are all wired together and act as one input on one channel. (non-standard wiring)

alt : Rogowski coil. : no simple interface, no currently available method to interface, little use in the DIY community. This seems attractive, but the support and availability is just not mature at this time.

I’m in a similar boat, but my need is to control Demand Use to bring down the bill.

studying the bills from the utility, they add complexity in order to confuse, and no matter what, there is almost no way to reverse calculate their figures.

I do maintenance work and work in most of the malls in the country. they range from a shared charge to sub metering with monthly meter reader to smart meters that feed back to the central station.

inevitably, you will get someone who complains and challenges.

I would suggest you use current rates, with say 5% mark-up to cover common use, outside lights, equipment and use, line charges, delivery charges, or whatever, same as the power utility, then show a 3% cumulative error discount to allow for the possibility of sub-meter errors.
and offer that any tenant can add a smart meter and have that feed back to your office for remote reading.

that way, if someone wants to add their own equipment to verify, then the accuracy of theirs and yours should be much less than 3%, so they benefit and there is no ‘error’ they can claim for rebates.