Ground mounted solar PV installation project

[minimutly]

Seeing as it persisted it down all day around here, I've got the current monitor pi up and running, just the one sensor at the moment.

Here's one line of serial output, only the figures in bold mean anything until I get the other two sensors up and running. This sensor is around the main incomer from the meter.

11 4.34 30.75 0.13 234.76 0.130 5.78 17.32 0.07 235.03 0.289 725.83 997.07 4.25 234.65 0.727

725.83 is real power in watts
997.07 is apparent power in watts
4.25 is the current being consumed in amps
234.65 is the RMS voltage
0.727 is the power factor.

so, I should be able to compare real power with what the meter is reporting, and it should be similar.

With the other two sensors I should be able to track down where those 4 amps are being consumed.

I'm waiting on some M10 channel nuts and some M10 Allen head bolts, so if they don't come tomorrow or if its chucking down again, I'll get the python tools installed and get it sending data into influxdb. I want to get some data in there before I turn on the solar.

Hmmm, something wrong there surely - what have you got that would throw your power factor that far off?
Would need something big with windings in it - like a lathe....

 

[brewdexta]

Hmmm, something wrong there surely - what have you got that would throw your power factor that far off?
Would need something big with windings in it - like a lathe....

I was pondering that, I can't think of anything inductive off hand. It was a quick snapshot, so maybe there was something kicking off that caused it. I'll leave it going a bit more tomorrow.

 

[brewdexta]

Looking at the output, the PF gets closers to 1 the higher the current drawn is. I rigged up another sensor and put it on the main incomer neutral, in theory it should be almost identical to live, any difference will be based on the slight difference in measurement time and any tolerances in the current sensor. Both sensors are the same model.

The following readings were selected at different times of the day when the load was different to highlight this

Blue is neutral, red is live. I put them is ascending order.

11 639.55 838.79 3.54 236.99 0.762 6.05 18.04 0.08 237.08 0.333 655.27 854.02 3.60 236.95 0.767
11 741.76 926.43 3.90 237.64 0.800 5.69 17.13 0.07 237.52 0.292 769.27 955.22 4.02 237.56 0.805
11 2216.59 2299.12 9.69 237.23 0.964 5.83 17.67 0.07 237.08 0.283 2255.08 2335.77 9.84 237.36 0.965
11 2269.07 2353.90 9.92 237.38 0.964 6.58 18.77 0.08 237.24 0.320 2296.25 2379.47 10.03 237.26 0.965
11 2947.38 3012.51 12.70 237.16 0.978 7.11 19.11 0.08 237.11 0.366 2916.22 2977.02 12.56 237.02 0.980
11 4010.47 4056.92 17.13 236.80 0.988 6.03 17.44 0.07 236.54 0.344 4252.25 4289.37 18.13 236.65 0.991

So I'm wondering if I have an inductive base load so at low amps this skews the power factor?

I can't think of any, I may turn off various MCBs to find what is causing it.

 

[minimutly]

Looking at the output, the PF gets closers to 1 the higher the current drawn is. I rigged up another sensor and put it on the main incomer neutral, in theory it should be almost identical to live, any difference will be based on the slight difference in measurement time and any tolerances in the current sensor. Both sensors are the same model.

The following readings were selected at different times of the day when the load was different to highlight this

Blue is neutral, red is live. I put them is ascending order.

11 639.55 838.79 3.54 236.99 0.762 6.05 18.04 0.08 237.08 0.333 655.27 854.02 3.60 236.95 0.767
11 741.76 926.43 3.90 237.64 0.800 5.69 17.13 0.07 237.52 0.292 769.27 955.22 4.02 237.56 0.805
11 2216.59 2299.12 9.69 237.23 0.964 5.83 17.67 0.07 237.08 0.283 2255.08 2335.77 9.84 237.36 0.965
11 2269.07 2353.90 9.92 237.38 0.964 6.58 18.77 0.08 237.24 0.320 2296.25 2379.47 10.03 237.26 0.965
11 2947.38 3012.51 12.70 237.16 0.978 7.11 19.11 0.08 237.11 0.366 2916.22 2977.02 12.56 237.02 0.980
11 4010.47 4056.92 17.13 236.80 0.988 6.03 17.44 0.07 236.54 0.344 4252.25 4289.37 18.13 236.65 0.991

So I'm wondering if I have an inductive base load so at low amps this skews the power factor?

I can't think of any, I may turn off various MCBs to find what is causing it.

I've had this when generating on a site comes very close to usage - the icoming current draw is tiny, so the power factor correction doesn't kick in - admittedly on a treatment works there will be loads or even mostly inductive loads. Anyway, even if the power factor is poor, at tiny currents it doesn't matter.

 

[brewdexta]

I've been feeling carp recently, the whole family seem to be taking turns to be ill, but I've still managed to get quite a bit done, the end is in site.

At the weekend I managed to get the worst job out of the way. This is the two lengths of 4mm hi tuff cable and two cat5e cables between the equipment room and the consumer unit via an existing hole in the gable end and two very small lofts, then outside under the eaves above the meter box. I just need to drill a 50mm hole through 18" of stone and it's done. The cable gets from the gable end below.

To the equipment room via a catenary wire that I also installed at the weekend. Here's the pole it terminates on next to the equipment room. There's also a wire from the pole to a post of the solar panel frame for more support.

The cables now hang from the catenary wire, but any photos wouldn't be for the faint hearted so I didn't take any of the process. Each end was ok, but the bit in the middle over a steep banking required the use of a neighbour, a plank ( in addition to me ), a bobcat, a ladder, rope, some unistrut, beam clamps and balls of steel.

It also required some nice strong wide cable ties. I could only find some locally that were 2 foot long, and only 10 in a pack so bought two packs. The shop owner said "she won't move much after all them!" as I was buying them. I suspect he didn't mean the cables

Other stuff is pre-cabling most of the inverter before fixing it to the wall in the equipment room. Its easier on the bench than it is when the underneath of the inverter is a foot off the floor.

I've also started wiring the PV fuses and isolators, and scrounged some cable tray from a neighbour to keep it all tidy.

The trend for solar cabling seems to be to use a length of trunking and tens of cable glands on the trunking.

I don't like that idea, I like to see the cabling. More pics of that when it's done, I can put the mains cables and cat5 into the room tomorrow.

Cheers
Andy

 

[Blobber]

What you want for catenary wire is these, you wrap the leather bit around the cable and tuck it through the slot, the bent wire bit goes through the hole and hooks onto the wire, it can all be done from one end and just slide along the wire

Catenary Cable Suspenders

Cable Suspenders are manufactured in leather with a hot dipped galvanised mild steel hook and brass eyelet. Used to support cables on catenary wire system.<br/> **Standards:** BS EN ISO 1461<br/> **Ma

www.remora.net

 

[brewdexta]

Interesting, I've seen them before, too late now, all done but I shall remember for the next time.

 

[MoreWellie]

I think this may be a new year upgrade

I have noticed this same compulsion in my son who I helped install a smaller roof system earlier this year and he has already upgraded it with batteries and an immersion heater dump

The inverter gives statistics

The statistics from the various devices don't seem to necessarily be as accurate as you might like, my son has quite different figures coming out of the growatt inverter, the generation meter and his supply meter

Have you used good quality screened cable for the cable clamps? I say this as they are very prone to picking up interference/fields from other cables if not. The screen should also only be earthed at one end of the cable

 

[brewdexta]

I have noticed this same compulsion in my son who I helped install a smaller roof system earlier this year and he has already upgraded it with batteries and an immersion heater dump

The statistics from the various devices don't seem to necessarily be as accurate as you might like, my son has quite different figures coming out of the growatt inverter, the generation meter and his supply meter

Have you used good quality screened cable for the cable clamps? I say this as they are very prone to picking up interference/fields from other cables if not. The screen should also only be earthed at one end of the cable

The main bit of information I should see is the export meter kwh light flashing less, and this bypasses the inverter and cable clamps. With that I can ascertain what we are saving based on previous history.

I have used screened cable for the CT so hopefully should results be reasonable. Interestingly the manual recommends extending it with standard cat5.....

Rather than use the cloud based web interface on the inverter, I will hopefully be polling the inverter stats directly using MQTT as per this link https://github.com/cmcgerty/Sofar2mqtt

I have all the components bar the buck converter I need to take PoE from the Mikrotik access point in the solar shed and convert it to 5v. I'm just looking at them now.

I have no idea if this will be better, worse or same as the stats on the web, we shall see

 

[MoreWellie]

the manual recommends extending it with standard cat5.....

Pretty sure that is just so less technical people don't have to make their own cables up

If the CT cables don't run close to any other power cables it isn't so important but I would stll used screened cables anyway

 

[brewdexta]

The weather forecast is carp for the next few days, but today was fine so got 8 of the 12 panels mounted. I'm hoping there's a couple of hours of decent weather tomorrow to finish the job.

A couple of the stainless bolts galled and slowed me down a bit. I did give them a spray of lube but could have missed some, I probably need to get some proper anti gall paste at some point.

The plank is bolted to the unistrut to stop the panels from sliding while I'm bolting them on.

Just got to set the angle on the remaining two bays then mount the panels. A few earth cables, then link the panel's PV cables and were nearly there.

Doesn't look too incongruous from the road. Quite a few neighbours are asking about installing their own panels.

The spark is coming around on Friday so better get a move on.

 

[Robotstar5]

Good job your in the middle of nowhere, kids round here would be chucking bricks at them

 

[addjunkie]

Good job your in the middle of nowhere, kids round here would be chucking bricks at them

Funny that was my first though too!

 

[brewdexta]

It had crossed my mind, there are a few feral kids about. The panels are more likely to get covered in takeaway wrappers from the gormless twits that jettison them out of there car window as soon as they finish filling their face. I may position a security camera near them, not that it will do much.

 

[Luke]

Seeing as it persisted it down all day around here, I've got the current monitor pi up and running, just the one sensor at the moment.

... I'll get the python tools installed and get it sending data into influxdb. I want to get some data in there before I turn on the solar.

I pull the data from the PV, car, and house in to Home Assistant. They've done a lot of work recently on energy monitoring, and in general it's a useful platform to have.

 

[brewdexta]

I've stayed away from all frameworks like Home Assistant, OpenHAB, Domoticz, Node Red etc. as I prefer to write my own stuff and be master of my own data. Once you have the data in a time series database, the rest is just presentation using Elasticsearch or Grafana or whatever floats your boat. But I am a bit of a nerd.

But I may have a play with Home Assistant at some point, it may be a good way to display stuff quickly to the family as I'm always short of tinkering time.

 

[knighty]

bit of a geek too, I gave Home Assistant a try, wasn't sure at first but I like it now

so quick and easy, plus most stuff / interfaces etc you can code yourself too

remote access / displays / checking from your phone etc all setup and ready to go


for solar etc. a tablet in the hall showing production/consumption/battery level etc. would be cool?

 

[Luke]

I haven't been around on here for a long time, and was only lurking for a while before then, but as I was writing my post, I guessed you might reply as you did .

I don't have the time to learn the skills I'd need to take your approach, Home Assistant has been a success for me in that it's achieved an acceptable end result without the investment of building from the ground up, and I've learnt a great deal along the way.

HA can be configured to work with an external database, and can play nicely with Grafana and InfluxDB. There's a heavy emphasis on data ownership and control, though I also subscribe to their cloud-based access platform, hosted be Nabu Casa, as it's less hassle than sorting it myself for off-site access to the app.

There are plenty of hardware and software integrations, I've used a lot of Shelly kit for energy monitoring.

And most importantly, it passes the wife test.

 

[brewdexta]

All panels up, connected and both strings are showing 218v within 0.3v which is about right so I'll call that a win. Earthing and cable tidy next.

The dog and I are off to the pub to celebrate.

 

[brewdexta]

The spark dropped by early today. I thought he was just coming for a look but he ended up connecting all the AC up, isolators and MCB so ready to go nearly. He's coming back Monday when I'll have the rest done for commissioning.

He also placed the a couple of additional current sensors in the consumer unit on the barn and PV circuits.

So this weekend I have to add the door to the equipment room, finish a bit of cabling and move the battery into there, finish of the earthing and add all the safety stickers.

In addition to that, I'm swapping out 2 old freezers for a new more efficient freezer. My base load without cooker, washing machine etc. is about 650W. I'm hoping this reduces it towards 400W. I now have the monitoring in place to tell, and the freezer arrives tomorrow.