Bringing my new system on-line

MichaelK · October 2017

For several years, my meager 400 watt solar system supplied a bit of power at my cabin. But, a long-term goal of mine was to design and build a much larger system capable of powering my 240VAC well pump. This would reduce fuel consumption on the homestead significantly.

Using a clamp meter I determined that the 1.5hp pump was using 10 amps to run. This Fluke meter also has "in-rush" capability, and measured that starting amperage at 37-38 amps (500 mSec). That means I need a substantial inverter to supply that much power. I selected Schneider's XM6848 120/240VAC inverter you see on the wall here. It has up to 12kw of surge (60 seconds). I still have to plaster up and repaint the wall where the solar electronics got mounted. I ain't gonna leave it like that.

I coupled this to 15-300watt Renogy panels, wired three in series per array, a 3S5P configuration. The 115VDC from each array was funneled through a combiner box which feeds a Midnight 200 charge controller. The Midnight charges a bank of 8 L-16 batteries wired in series for 48V. Here's a pic of the array frame, made out of welded unistruts.

The frame fits over a single steel pipe sunk in 48" of concrete, which allows the array to track the sun left and right. The horizontal stabilizers at the bottom also allow for season angle changes. I call it "Hillbilly solar tracking". Here are the arrays with the panels installed..

Though I did most of the wiring myself, my BIL is a retired electrician, so I had him inspect my work. I'll be sealing up the plastic conduit and getting it buried before winter comes.

Last month we flipped the breaker to start the well pump, and the XM powered it up with ease. I'm now in the performance qualification phase of the system. My rule of thumb is to run the pump only while the panels are capable of making >2400 watts of power. This happens after about 9:30 in the morning. If the arrays are swung over to face SE, the batteries are mostly charged by 9am. By noon though, there's enough excess power to bring the batteries up to float. I can continue pumping till about 3:30 in the afternoon if the arrays are swung SW. With the new solar power, I've gotten my 5000 gallon tank up on the hill completely full to the brim!

I have a video of the pump running, but that's too big a file to post here.

Photowhit · October 2017

Nice!

While it's hillbilly tracking, you might also consider stetting it up for virtual tracking, with 2 or 3 facing south east to charge up after the night time loads and 2 facing south west so that your afternoon early evening loads run off the array instead of batteries.

MichaelK · October 2017

Yes, I'm thinking that too. I can orient the two eastmost arrays to the SE, while setting the two westmost arrays SW. That would tend to reduce the peak noon power, but spread out the time I'm getting at least a kwh of power to between 8am and 4pm. That's what I have in mind for off days when I'm not pumping water. On irrigation days however, pumping takes priority over everything else, and plan on tracking.

One thing I'm starting to find is that when oriented SW, I start to get shading after 4pm, and a due south orientation actually produces a little more power. If you notice the second picture, I positioned the arrays on a slight incline to give an advantage to early morning charging, at the expense of late afternoon.

Ultimately, I plan on installing another array that I already have panels for (the YingLi's I mentioned in another post) that will be optimized for due West production. That can throw out another kw to play with in the late afternoon

westbranch · October 2017

the secret is in getti9njg as many hours of input, not just the max power at a set time of day,say noon.... It may require a bit of tinkering ... which panels get the sun longest in the PM, point those West and the others S to SE

MichaelK · October 2017

A couple of observations. Historically, I've kept detailed records as to hours of generator use, fuel consumption, and number of gallons of water pump. A long-term average when powering the well-pump via the generator is 275-280 gallons of water per hour. Now that I'm switching to solar one of the first differences I'm noticing is that gallons per hour has gone up to about 320-325. What this suggests to me is that the XM inverter is producing higher quality AC than the generator does. What I've seen in the past is that under generator power the draw was 10.0-10.1 amps at 240VAC. With the XM I'm seeing 9.6-9.7 amps at 235VAC. I'm assuming that cleaner power is being produced by the XM, and the pump motor is running more smoothly?

A second observation is in regards to using the generator for charging. I purposely drained the batteries somewhat so I could test generator charging. The default charging setting for the XM is 140A, so I reduced that to 27% to match C/10 for my Trojans (370/10=37A). The XM would not qualify the generator at that setting. I had to go down to 20%, which works out to be only 28amps. 28A X 58V=1624watts. Once the power got qualified by the XM I then tried to inch up the charging by 1% each minute or so. That worked, and over several minutes I got up to 37amps qualified. But, when I inched up to 40amps, the XM dropped the generator's power again. After that, I couldn't even drop the percentage back down to 20% and get it qualified. I had to physically shut off the XM and restart it before it would quality any power from the generator at all.

So it seems like my generator can't handle producing more than 37A X 58V= 2150 watts cleanly. It's an AC-Delco 6000 watt generator with about 900 hours on it. I'm guessing that I've just about worn it out. Would this warrant purchasing one of those little hand-held oscilloscopes to check the quality of the power? My idea would be to test the generator's output while putting simple induction loads like the toaster and toaster oven on it. Opinions?

BB. · October 2017

It is difficult to tell why the difference between the genst and XW without doing more analysis... It simply could be the difference between 240 VAC (genset) vs 235 VAC (XW).

Your toaster is a "resistive load". Your pump is an inductive load. Gensets (and inverters) can have some pretty nasty responses to different types of loads. And oscilloscope can show the more obvious differences (very poor wave forms). But to really analyze the differences, you need to measure both the amplitude and the phase differences between voltage and current wave forms. For smaller differences, you would have have difficulty seeing the variations in the wave forms on a standard scope screen.

-Bill

Estragon · October 2017

A couple of thoughts:

- It may be worth considering the length, size, and condition of the generator wiring, check for a loose connection, etc. Higher current means more wire losses, maybe enough to drop voltage out of spec? May also explain higher draw for the pump vs XM?

- battery chargers (and many pumps, for that matter) can have pretty poor power factors, so the generator may have to produce more to do 2150w worth of battery charging "work".

MichaelK · October 2017

Estragon, that's a great idea. I was thinking along the same lines to open up the generator enbel and front panel and detach and shine each of the electrical contacts. The brass may have corroded over the years, and high resistance might be the explaination for what's I'm observing.

Estragon · October 2017

Might be worth checking downstream output wiring, retorque breakers, etc. too.

Bringing my new system on-line

Comments

Categories