OpenEnergyMonitor Community

A more industrial/commercial angle

Hello folks,

I work with an energy management company in New Zealand. I deal with many public institutions that are starved for funds yet very inefficient. The majority of the people I interface with would love to have sub-metering, but the up front costs are too high and they cannot get the case past the CFOs. They do not need billing accuracy either, so they do not need the most expensive of equipment. They just need inexpensive diagnostic tools to provide finer data granularity.

I can see that the IoTaWatt could be a great tool for this but I am trying to clear up a few issues in regards to the limitations of the equipment. I am wondering if folks here can help me out a bit?

I can see that the IoTaWatt can handle CTs of 200 amps and I am wondering if and how it would handle larger CTs of say 500 amps? Is it possible and how would I make this work?

Also, I have seen that there is a discussion in regard to 3 phase systems. I can imagine the IoTaWatt being used to monitor pumping units and chillers in a three phase environment. Is three phase something that is well and truly on the way or is it difficult to implement with the device?

Furthermore, I see that power factor is being reported by the software. I imagine the device is taking a voltage reference from the power supply for the unit. That would mean that the unit is providing an accurate power factor for one phase, and using that same voltage to calculate an approximate PF for the other two phases if used in a three phase environment. Is this the way it works? If so and if people want to use the units for true three phase environments, it may be worth investigating creating another probe to measure a secondary voltage, so that two references are measured and one is calculated, thereby making all three PFs accurate. This would allow for measurement of a 3 phase/3 wire system with 2 CTs and 2 voltage probes.

Last question- Is there a way to look at any real world data as presented in the graphical interface? I am curious as to what graphs are available and how flexible it is.

Kind regards,

1 Like

Hello Pat,

I’m the developer of IotaWatt. You ask a lot of questions about capabilities and potential of features of the device that, while considered in the design, have not yet been exploited, tested and documented. I’ll try to give you a sense of what’s possible now and what’s possible going forward.

While there is no independent data available on accuracy, I’ve got monitoring gear alongside some revenue grade meters and monitors I have high confidence that it’s at least 2% and high hopes it will turn out to be 1%. That said, there are no guarantees at this point.

Short answer - it should work and there are a few different ways to approach it.

I can’t see any reason why CTs with higher current-ratios can’t be used. It’s a matter of sourcing the appropriate CT for an application. For 500 Amps maximum to connect directly as delivered, you would need a CT with a current ratio of 12000. As you get into these larger and costlier CTs, it may make sense to use an IotaWatt with no internal burden resistors and use voltage type CTs for more flexibility in the field. As an example, a WattCore 400A split core CT with 1Volt output is around $50 (US).

Three phase support was always part of the consideration in designing IotaWatt, but I’m not a power engineer or even an electrician. I’ve concentrated on the residential application for testing because that’s the immediate user group and because I have ready access to residential test opportunities.

Short answer - three phase capability is there now but untested and not documented. I’d work with you if you are interested enough to be in the vanguard.

The product is far enough along, and there is capability designed in but not tested for three phase. It just needs a user community to try it out, and work with me to work out an enhanced user interface for the various three phase approaches. There are a ew different ways to do three phase, basically trading off accuracy for simplicity. You won’t know where you fall on that curve until we know more about the accuracy of the various simpler methods.

For example, if you have a few different large motors on compressors and pumps and, as you have indicated, you are interested in ballpark numbers as to what each uses, it might be accurate enough to just measure one of the legs and multiply by three. There is a lot of symmetry here and there’s are few reasons to think one phase will be radically different from the other two in a motor.

Short answer - IotaWatt can do it both ways. It can use multiple VTs, one for each phase, or it can phase shift one VT as a reference for the other two phases.

I’m assuming you are looking at power factors on the mains as opposed to a particular three phase device like a motor. If power factor for each phase is a concern, using three different voltage references is probably the way to go. Also, power factor is not kept in the datalog, although there is space for it. It is currently only available in the real-time status report. There is a rework of the datalog in progress that will record frequency on voltage channels and power factor on power channels. How that will be reported out is still unresolved.

Short answer - buy one.

The local graphical interface is an adaptation of the Emoncms graph program. There are some subtle differences, but the concept is the same. Also, given that you can upload the data to a local instance of Emoncms, all of the graphical tools of Emoncms are available.

Another option in an industrial environment might be to upload the data to an influxDB/Grafana server. You can build your own, there are a few lower cost services around, and there is a somewhat pricey but extensive capability hosted on AWS.

Hope to talk to you soon.

1 Like

Hi Bob
I’m happy to work with you on the three phase development, I have IotaWatt ready to be installed.
Is there any value to sort you out with VPN access to it to aid in the development so you can tinker remotely?
The installation environment is a UK office with LED lighting, computers and air con with a 100a three phase supply.


Thank you for the detailed answer Bob. I will digest it all and get back to you on it.

OK. This week is pretty well booked up for me, so it’s best if you go ahead and install it, using the Wiki instructions, and familiarize yourself with the configuration app. If things get really messed up or really interesting, I may take you up on the VPN offer, but for now I’ll let you handle the local communication.

Once you have the CTs on the mains and any other circuits, we can do a chat and go through the configuration settings to phase shift the CTs over the VT.

Any questions, fire away.

Do you want me to use one VT or three?


Three would be terrific. I have every reason to believe that with three, accuracy will be on a par with single phase .

Once the three VT baseline is established, I’d like to add another set of CTs to the mains and monitor them simultaneously using the phase shifted CT referencing just one of the VTs. That will provide a lot of insight into the relative accuracy of the simpler phase shifted method against the gold-standard of three CTs.

No worries, I’ll have a look for some cheap DIN mounted 12v VTs to ease wiring them in as I believe you can calibrate for non supplied VTs.


Yes you can, and IotaWatt accepts a 12V AC VT, so that’s good. On two that you feed into the input jacks, try using something around 350-400 ohms (1/2watt) in series instead of the 288 I recommend for a 9V adapter. That should put it in the range of the 9V with respect to current and lower the power dissipation. Use the plain generic with a starting cal of around 20 and calibrate. The only wild card is the phase shift. If they are anywhere near the weight of the Ideal adapters, 2.0 would be a good guess.

Cool, I’ll have a look online for appropriate VTs and get them ordered.


“or it can phase shift one VT as a reference for the other two phases.”

Is there any information on how to do this ? I could not find anything on 3 phase configuration in the wiki. I think I understand how to setup with three voltage sources (one for each phase) but not this other approximation method.


There is nothing published yet. As soon as some of these discrete VT type sites get up and running, I’ll try to get a few to set up simultaneous monitoring using both methods to get some data regarding accuracy vs the true voltage/phase reference. This can be done on the same device - one CT monitoring the circuit with a discrete voltage/phase reference, and another with a phase shifted reference from a different phase.

I have done it using my US split phase system where the other phase is 180degrees, but the voltages are related and the phase is exactly 180 degrees. But as a result I know the IoTaWatt phase shift algorithms are working, so it’s just a matter of degree - so to speak.

If the method proves to be accurate enough to be of interest to users, then I would enhance the user interface to support configuration. In the meantime, the way it can be done with the existing utility is to configure the CTs as “generic” and add 120 or 240 to the phase parameter.

For instance, when you configure an SCT013-000 CT, it has an implied turns of 2000 and a phase of 3. So if you were to install it on a phase that is shifted 120 degrees from the primary reference VT, you would configure it as:
The phase correction software would then slide the voltage samples an extra 120 degrees under the current samples.

Thanks Bob,

I did consider using the CT phase shift to create an approximate for the other two phases but was not sure if the algorithm would work - looks like it should do I will give that a try.

Is current logged or just watts ? For a simple load survey job then amps is accurate enough and I don’t really need pf or watts and I know what circuits are on what phase.

Excellent project, I was starting to look at a similar concept but you have done all the hard work. Ordered my iotawatt last week so hoping to try it out soon.


The log contains only watts. Current and PF are available at sample time, but not recorded in the log.

PF and Irms is reported out damped in real-time via the status API.

Hello @overeasy . I work in a High voltage power sub-station where we collect time series data manually from energy meters that measure as much as 300 megawatts of energy.

  1. Can IOTAWATT be used to collect this data directly from the high voltage switchyard where we have 1200:1 turns ratio CTs with maximum output current of 1A installed on 330KV/132KV/33KV power line equipment?.

3.Can the data on the EmonCMS cloud be integrated with third party IOT platforms like google sheets, thingworx, siemens mindsphere, ge predix, Ubidots etc?

Hi Don,

I would be doing you a disservice in offering advice in this environment. I’m completely unqualified to appreciate the hazards that may be present.

Thanks for the interest.

You would need to use an interposing 1 A : 50 mA transformer for this. The problem here is, those are rare beasts - you might need to have one custom-wound or, at 1 A secondary current from the existing c.t, you could very easily use a multi-turn primary winding in the interposing c.t and use (say) a 20 A : 50 mA one, which might be easier to procure.

(Would I be right to assume that the 1200 : 1 c.t. is feeding other instruments too?)

That would sort out the electrical side, but Overeasy would need to say whether his gear could be calibrated for what would effectively be a 1200 A : 50 mA c.t. and two phase errors. Provided that one side of the 1200 A c.t. secondary is earthed, there shouldn’t be a problem.

Then perhaps @TrystanLea can answer the part about sending the data to the other places.

There is no in-built integration with these platforms, but emoncms does have a HTTP api and so you might in theory be able to link the two…