Monitoring of two buildings with solar in North America

Robert.Wall · 2020-03-05T09:29:04.081Z

fuzionworx:

I want to be able to collect data sufficiently granular enough to characterize the loads, sources and loses across the entire system independent of any vendor specific component.

I did think that could be a long-term aim.

fuzionworx:

There may be value in having these to account for feeder loses in the future.

I don’t understand - you know the current, there’s almost no loss to capacitance, so it’s the same current at both ends of the cable (certainly to better than the accuracy of the c.t’s). The losses are almost all I²R and you can calculate that.

fuzionworx:

I imagine that CT’s could be clipped around multiple conductors, but there are a mix of 120V and 240V circuits.

That - depending on how much spare wire you have - should certainly be possible with the 10 circuits handled by the transfer switches, one c.t. per leg (automatically accounts for 120 V & 240 V circuits if you use 120 V as the voltage reference.

fuzionworx:

do you mean that we should also be measuring voltage at the inverters?

Yes - or at a circuit fed by the inverters. When the mains disappears, you’ve got no voltage reference for those loads.

Bill.Thomson · 2020-03-05T17:02:25.211Z

fuzionworx:

I imagine that CT’s could be clipped around multiple conductors,

The main caveat there is to be aware that circuits on even numbered breakers are on one
hot leg and circuits on odd numbered breakers are on the other hot leg. So you’ll likely need two CTs to read all 10 circuits.

That said, with only 10 circuits, it is possible all 10 could be on one leg.

Robert.Wall · 2020-03-05T17:38:56.581Z

All Daniel needs to do there is imagine a pile sheets of this drawing, but each page having only one load, with their appropriate line conductors all passing through one or both c.t’s:

https://learn.openenergymonitor.org/electricity-monitoring/ac-power-theory/files/N_America_Fig_2.svg

The important thing is not to get the A and B legs mixed up.

Bill.Thomson · 2020-03-05T17:48:56.208Z

Robert.Wall:

All Daniel needs to do…

It’s simpler than that.
All he’s got to do is pass wires attached to even numbered breakers through one CT and do the same for the wires attached to odd-numbered breakers.

Robert.Wall:

The important thing is not to get the A and B legs mixed up.

Keeping track of even and odd numbered breakers takes care of that.

fuzionworx · 2020-03-05T18:01:34.191Z

Robert and Bill, thanks for the feedback, I really appreciate it. The lines from the transfer switch do alternate between the 2 phases, and any two adjacent lines can be used to create a 240V circuit. After reading your posts, it is obvious to me now that I will need 2 CT’s that each should contain the 5 lines from the same phase, Doah!

Bill.Thomson · 2020-03-05T18:06:54.091Z

Not a problem. I’m in Oklahoma, so am familiar with US residential electrical service.

jdeglavina · 2020-03-12T10:16:15.965Z

If you need CTs that will fit the 4/0 mains, I have these 16mm ones in stock.

fuzionworx · 2020-03-13T04:13:45.501Z

Thanks John, I will check them out.

Bill.Thomson · 2020-03-13T11:36:33.137Z

jdeglavina:

that will fit the 4/0 mains,

I had to trim the plastic on the movable part of the SCT-016’s shell to get them to fit my SEWs.

jdeglavina · 2020-03-13T12:00:47.571Z

I guess it depends how thick the insulation is on your mains. 4/0 is 11.68mm + insulation thickness. From what I can tell the most common thickness is 1.27mm with a jacket of 0.18mm in the US in accordance with UL-1083

Bill.Thomson · 2020-03-13T12:11:48.235Z

It can vary quite a bit. e.g. anywhere from 0.624 to 0.635 inches for THHN AWG 4/0 wire.

image522×531 15.3 KB

awcwire.com

4/0 AWG THHN Building Wire | Allied Wire & Cable

image835×605 65 KB

usawire-cable.com

thhn_thwn.pdf

82.91 KB

fuzionworx · 2020-03-14T19:58:23.741Z

Thanks for the information Bill and John. Sorry for the slow response, this week has been busy with my day job and everything else that is going on. Here is a photo of the Shop Main Panel and transfer switch (a work in progress):

shop_main_panel202003142268×4032 1.87 MB

The Service Entrance cable is 4/0 URD, the outside diameter is .66 inches/16.75 mm. So I will probably have to look around for something else. I have been working first on the emonCMS server and emonTX/Ethernet shields/Arduino integration. But will need to dive into the sensor sourcing in the coming week.

stephen · 2020-03-14T21:17:00.874Z

quite a bit loose wire that is neither in the wall , stabled down or in conduit- renovating ??

Bill.Thomson · 2020-03-14T21:35:48.382Z

NOT a problem, Dan. Sometimes “life” just gets in the way.
It looks like you’ve got plenty of room above the main lugs for CTs with a 3/4 inch wire window.

Typically, that family of CTs is the voltage output type. (333 mV) not the preferred choice
for use with OEM hardware. Variants with current output are available, you just have to do a little
searching. Here’s an example:

amazon.com

WATTCORE WC3 Split-core Current Transformer 100:0.1A, Opening ID 0.75"x0.96":...

WATTCORE WC3 Split-core Current Transformer provides precision measurement for sub-metering, controls and energy management without interrupting the power line. Ergonomically designed with secure tabs, this CT is ideal for...

They’re not outrageously expensive, but they’re not low cost like the SCT-013 and SCT-016.
Cost is usually 40 to 50 bucks each.

fuzionworx · 2020-03-14T23:49:29.570Z

Yep, like I said, a work in progress. There used to be just 100A service with smaller service entrance cables (2AWG). I upgraded to 200A service with new panel and larger mains (4/0) to reduce voltage drop over the 160ft/50M run to meet code for grid tie in the future. This was part of a major expansion of the building that also relocated the panel adjacent to where the solar equipment is housed. I am currently working on extending metal conduit for all the circuits.

solar_equipment_202003144032×2268 1.94 MB

Bill.Thomson · 2020-03-14T23:50:14.858Z

Looks good!

fuzionworx · 2020-03-30T05:53:24.564Z

I have made a some good progress on the wiring in the shop, and am able to get back to the monitoring projects. I have been looking at possible CT sources for the 4/0 AL URD conductors, and have found an affordable option on eBay from US sellers. There are several listings for these SunPower PVS5 monitoring kits with a pair of Jiangyin XH-SCT-0750-100 CT’s for $20-$30 USD, including shipping. e.g.:
Sunpower PVS5 Outdoor Metering Consumption Kit 522398 Jiangyin XH-SCT-0750-100
@Robert.Wall mentions in Help with whole house power monitoring (200amp/USA) that one should replace the SMT burden resistors on an emonPi with 33Ω versions for similarly spec’ed CTs. I assume I should need to do a similar mod to emonTX shield (honestly I haven’t looked at the boards yet). I would like to order the CT’s early this week, and start diving into the boards later on.

Bill.Thomson · 2020-03-30T08:33:06.826Z

Unfortunately, those CTs are the voltage output type.

What you want is an unburdened CT that puts out current vice a voltage.
IOW, for what you’re wanting to do, it’d be a waste of $$$ to order them.

Here’s an example of what you want:
https://www.ebay.com/itm/153648731718

and here’s another:
https://www.ebay.com/itm/Magnelab-SCT-0750-000-000-Amp-Adjustable-Current-Transformer-Split-Core/254415198366?_trkparms=aid%3D1110006%26algo%3DHOMESPLICE.SIM%26ao%3D1%26asc%3D225074%26meid%3D20d4fc8db4934df1a82bb2a3e896f10f%26pid%3D100010%26rk%3D1%26rkt%3D12%26sd%3D153648731718%26itm%3D254415198366%26pmt%3D0%26noa%3D1%26pg%3D2047675%26algv%3DDefaultOrganicWithImageAndTitleNsfwFilters&_trksid=p2047675.c100010.m2109

Robert.Wall · 2020-03-30T10:49:36.346Z

The plus side of a voltage-output c.t. is that the output voltage is accurately calibrated by the manufacturer. The minus side of a voltage-output c.t. is that the output voltage is accurately calibrated by the manufacturer and you can’t change it (upwards) to suit your individual circumstances. You can of course add a second parallel burden to reduce the voltage, but that’s hardly ever necessary.

In your case, you need to choose the burden resistor so that at the maximum current you’re interested in, the burden develops no more than 1.6 V. (I’m talking rms values of course.) The equivalent voltage for the emonTx and emonPi is 1.1 V rms, because the ADCs have different voltage input ranges.

fuzionworx · 2020-03-30T15:33:42.992Z

Thank @Bill.Thomson and @Robert.Wall, as always your expert feedback is appreciated! I was questioning that in the back of my mind, am glad I posted this before proceeding. Will update with status later this week.