Managing and using your generated power - a whacky idea?

[ProDave]

A1) My solar PV is 1.9KVA peak. The object is to use as much as possible of the generated power rather than let it be exported. the "background" usage, i.e everything on standby, fridge, freezer etc is about 400W, so I rarely have much more than 1KVA going "spare" even in full sunshine. So switching on a 3KVA load, would mean I was importing (i.e paying for) 2KVA of that.

So by using a low power load, I can use more of my free, self generated power, without importing and paying for any.

A2) read the thread on the other forum about the basic design I am now using. As well as taking a reading from the current transformer, it also samples the voltage from a transformer. By looking at the phase of the voltage, it can determine if the current is in phase (exporting power) or out of phase (importing power). your normal electricity meter must do something similar to ensure it does not meter exported power.

 

[OnOff]

ESR Electronics do lots of home automation kits and I/O stuff etc:

Velleman - Velbus Home Automation System

Also Carlo Gavazzi - do current monitoring kit:

Carlo Gavazzi - Product Selection

 

[ProDave]

^^

"The Velleman Velbus System is a 4-wire Home Automation System. Perfect for the DIY enthusiast or professional contractor, thanks to it's easy installation and very competitive price"

I don't thinks so, typically

 

[Adrian41uk]

As well as taking a reading from the current transformer, it also samples the voltage from a transformer. By looking at the phase of the voltage, it can determine if the current is in phase (exporting power) or out of phase (importing power). your normal electricity meter must do something similar to ensure it does not meter exported power.

Dave, I'm not sure this is correct. The phase angle between the voltage & the current is the power factor not the direction of power flow. There are methods to the direction of current flow using a polarising voltage but this isnt as simple as your method. Remember, for an inductive load, the current lags the voltage & for a capacitive load, the current leads the voltage.... for a resistive load they are in phase.

 

[kme]

^^

That was the way I understood it - which returns to the original question.

Even with a ct on the inverter cct, which can provide the current generation figure -

and another on the consumers (or meter) tails ( to provide an over-usage OR export reading)

I still haven`t got my head around how you can determine which is which; unless you ct`d every final cct, and summed the individual usages..........

I`m off for another beer -it might help the thought process.

n.b. Dave. Other thread where?

 

[SteveT]

Or am I just bonkers?

Probably!

 

[ProDave]

What's going on?

I posted a reply earlier this morning. Then went back and edited it. Now it's gone. Forum time warp?

I'll try again and see if this one sticks.

Re the method of determining current flow direction.( Link deleted).

If that doesn't work, then in my case I can use two CT's. My PV connects to it's own mini CU, so a CT in the tails for that would measure generated power, and a second CT in the tails of the main CU would measure used power. Then it's just simple arithmetic.

I'm still waiting got my arduino so development is halted until that arrives from Hong Kong. The CT ordered from Hong Kong at the same time arrived last week.

EDIT: If the previous post was deleted because you don't like the link, then just delete the link, and please let me know if I have broken any forum rules. Otherwise I am just left wondering what happened.

Admin deleted the last post Dave, I have removed the link so your post can stand.

Edit #2: Thanks for the clarification Admin.

Edit #3: I am lost (Ad3)

 

[ProDave]

Okay, My arduino has arrived and I have it wired up.

It took a couple of hours to get the arduino programming software to run on my PC. All to do with Windows XP and the problems it has dealing with USB to RS232 converters and the messing about with drivers you have to do to get it to work.

so now I have the hardware and the programming environment set up, I've started programming.

The method I am trying of determining current flow direction is not entirely working properly. It does to some extent give an indication of direction, but it's by no means perfect. I don't yet have it controlling the dump load directly, just turning an LED on and off while I get the software to work as I want it.

What's needed to proceed is to lug my (big by modern standards) oscilloscope down and set it up in my office where the PV dump controller and my PC are, and have a good look at the waveform coming out of the current transformer to determine if this method really will work or not. That of course also needs a nice sunny day so the inverter is generating nicely. so it could still be some time before I get it cracked, but progress is being made.

What I can say so far is I do like the arduino and it's programming environment. Easy to program (well as long as you are used to programming in c) and easy to download and test programs. Looking at the arduino website it's well supported with plenty of libraries and functions for just about anything you could care to mention. So highly recommended for a simple embedded controller for just about anything, and dirt cheap too at just under

 

[Adrian41uk]

Hi Dave,

I think, when you put the 'scope on the CT output, you'll see that the waveform is a very poor sinewave - its going to be destorted by harmonics generated by switchmode power supplies & other electronic devices in the home. I assume that you are driving the CT into a burden resistor & looking at the voltage developed across that? (Remember that one should never run a CT with its secondary open circuit or driving into a high resistance). Also, I suspect that the output of the inverter isn't a true sinewave ..... I suspect that its controlling the firing angle of its output thyristors to inject a pulse of current back into the mains just ahead of the peak of the mains sinewave.

I think that the other problem will muddy the waters is reactive power. Although you'll manage to get the real power to reverse I doubt that you'll reverse the direction of the reactive power flow. Since the current will contain both real & reactive components it may well be case that the apparent current doesnt actually reverse in direction.

My post last week prompted me to go find out how a commercial power meter works out the direction of power flow ..... I did find an answer but it isnt easy..... can you do floating point maths & trigonmetry on your ardunio?

Adrian

 

[ProDave]

My post last week prompted me to go find out how a commercial power meter works out the direction of power flow ..... I did find an answer but it isnt easy..... can you do floating point maths & trigonmetry on your ardunio?

Adrian

I would be VERY interested to see that information. The arduino can do floating point maths, and has basic sin. cos, tan, square root functions etc. At the moment I am doing everything with integer maths.

Anyway I have made a LOT of progress, in fact it's almost finished:

The problems I was having a couple of posts back were due to the zener diode across the CT. I removed this completely. It was there in the original design to protect from very large currents. Instead I have reduced the CT load resistor from 1K down to 150 ohms. This means I will get lower input values but it will cope with a much higher current without overloading the input, so removing the need for the zener diode in the first place.

I took the original idea and extended it. I now separately sum the current over the first and second half cycles to give a measure of imported current and exported current. I have tweaked some of the timing values. I use two spare debug digital outputs to show me the exact time periods the current is being sampled and using the oscilloscope fine tuned the timing to get a current measure in the centre of each half cycle. I also increased the number of samples per half cycle.

I've attached the code (as a text file) and some oscilloscope traces of the output from the current transformer. The top trace, the square wave is one of my debug timing signals, the bottom trace is the actual current.

For some reason the attachments appear in the opposite order to what I expected.

So the Last picture, shows importing a large load (3KW kettle on) Second from last picture, importing a small load, just all the stuff in standby, pc etc, panels not generating. Note the rough non sinusoidal waveform. third from last picture is as near as I could get to equilibrium, generated power = consumed power. the very rough waveform means the controller shows both a small imported and a small exported value at the same time. First picture exporting over 1KW. This clearly shows the phase reversal between importing and exporting that allows this method to determine the direction of current flow.

In spite of the rough waveform around equilibrium, it does seem to work and make a sufficiently accurate measurement of current and direction.

I'm still running this off line, just monitoring the readings. I now have enough confidence that I will close the loop very soon and connect the dump load for real. (but want to do that first on a day when I'm going to be in to keep an eye on it)

Then the next plan is to add some data logging to record how long the dump load has operated for.

PV_dump_v1.txt

 

[steptoe]

excellent work Dave,

looking good

 

[roys]

Good stuff Dave, looking forward to seeing the finished article.

 

[SLIPSHOD & SLAPDASH]

I haven't got a clue about 99.9% of this thread & we are not registered for PV installs either. However we did one very large job (where we did the wiring and some clever French guy watched us and programmed it plus 'signed it off') where the house has every form of generation, 3 heat loops under the drive, lawn & pond. Windmill, PV panels that are individual tiles that look like slates and cover the entire roof and solar/thermal water panels. All the electrics are generating 3 phase (i know each pv is single ) and these charge batteries. The battery room is about 6ft x 6ft. Its a large country house and they recon they can go off line for over 2 months. I suppose cost is the main reason people don't do this. Last time i was there some government officials were looking over the place in regards to the energy efficiency. The TV prog Grand Designs wanted to feature the house but the family want to keep it private. I have no pictures but have told my partner next time he is going down there i want to go with a good camera

 

[ProDave]

Interesting link Canoe.

There's actually quite a lot of "systems" out there.

I found this one on ebay eBay - The UK's Online Marketplace

PV_dump_v2.txt

 

[ProDave]

Results time:

Yesterday was a good day for testing that the dump controller works as it was mostly cloudy with a few sunny spells so was good for checking it turned on and off properly according to conditions.

Today was almost wall to wall sunshine. My 110V powered, 700W immersion heater clocked up 250 minutes of ON time today. That's not bad considering we also did 2 loads of the washing machine and tumble dryer, and one load of the dishwasher (all sequentially one machine at a time)

It's interesting that the dishwasher and washing machine only consume a lot of power when heating the water. When the heaters turn off, the dump load came on. Unfortunately that's not true for the tumble dryer that's a constant high power load. God I hate tumble dryers, but Mrs PD insists, "it's the only way to make the towels nice and fluffy"

So 250 minutes at 700W is roughly 3KWH of free power "recovered" Not enough to fully heat a tank of water, but certainly a saving of oil. It bodes well that on a less busy day without all the other machines going, there should be more spare power for water heating.

Another statistic, the normal usage meter clocked up just 1KWH in that time. That's not bad for 2 washing machine and tumble dryer loads, and a dishwasher load.

That also shows my hot water strategy is working. I still have the oil boiler coming on in the morning to heat the water, but it turns off at 7:30. Just after that Mrs PD likes to take a shower, and being a woman, takes a long time in the shower, so uses a lot of water. Previously the boiler would have fired up again to re heat it, but now it's left colder, to leave capacity for solar PV water heating.

 

[steptoe]

dave, this is a very interesting thread,

Im liking what you are doing, and developing,

 

[dbicket]

Top work ProDave:Salute. Still lucky enough to have old analogue meter. But will defo give something like this a shot when they change it...

 

[steptoe]

dbicket,

it makes absolutely no difference what sort of meter you have.

what are you thinking of?

 

[NozSpark]

I think his meter is going backwards

 

[dbicket]

Yea, it goes backwards! For now anyway. Probably just jinxed it. Utility company will be at door now Monday morning wanting to change it..