The COSF Solar Well


Over the last few weeks, I've been doing some testing of our "solar well", the large black tank near the pyramid. Herein I'll give an overview of the results. Take all estimates with a skeptical eye unless I specify exact measurements, which means I've verified such.

First, the electrical setup: 4 solar panels, nominally 12V each with a no-load maximum of about 22VDC, are hooked up in series and meet at a circuit breaker at the northeast corner of the structure. An earth ground also feeds into the same box, as does a separate ground wire (distinct from the negative lead) from the solar (photovoltaic) panels. When the switch is turned on, roughly 80 volts are fed to the SC-1PV Sunrise Submersible pump controller through a short length of what looks to be 3-conductor SJ cord with the green ground wire snipped. Out of the controller you'll see some thinner SJ cord, which goes to the float switch, and a flat 3- or 4-conductor cable which goes to the pump.

When the tank is full, the float switch activates, and goes from its normally "open" state to "closed", at which point the pump controller will refuse to pump. When I first checked the pump a few weeks ago, it did not turn on at all; when I opened the cover of the controller a day or two ago, I noticed a couple of strands of wire from one side of the float switch connection coming near, and possibly touching, the other side, which would explain that failure; a short would cause the controller to "think" the tank were always full, thus the pump would never be engaged.

In any case, after moving those errant strands of copper away from the other side of the connection, the pump attempted to turn on. Unfortunately, it was too late in the day, and though it showed a high level of charge with the LEDs, once the pump switched on it proved how weak the sunlight's oblique rays were, and it instantly shut off again. Today, however, the pump ran fine since about 11 AM, and sometime this afternoon (July 2, 2007) the float switch was lifted by the rising water and the controller shut the pump off. I got there after the fact, and verified that operation by running the water for a minute or two, draining possibly 20 gallons from the 1550-gallon tank. The float switch again deactivated, the pump turned on, and within 10 minutes or so the level again caused the float switch to activate, and the pump stopped.

The water flow, as I understand it: the pump sucks the water from 300 to 400 feet down, sending it up a 3/4 inch schedule 40 PVC pipe, through a gate valve about 4 feet above ground into the top of the tank. A 2-inch PVC pipe exits the tank, and has a faucet tap with 3/4 inch hose thread about 2 feet above ground; it then goes underground to a gate valve and meter visible in the westernmost of the two steel 50-gallon drums buried on the north end of the structure. Where it goes from there I have no clue, but the valve appears closed, and I was told it must remain closed because the pressure from the other well would cause backflow into the tank and overflow it.

Some other observations, of dubious value: the ground lead measures roughly half the charge of the solar panels, so the negative is truly negative with respect to ground, roughly -40VDC in a no-load condition and presumably -24 when the pump is engaged. It makes me want to consider re-doing my 12V solar power setup with the center grounded, thus giving me +6 and -6VDC in addition to the 12VDC. Also, there is a length of plastic tubing running from the bottom to the top of the tank, ostensibly to serve as a level indicator, but it is so opaque due to UV degradation and rust buildup that it is useless for that purpose.

This large PDF document gives cost estimates per gallon for similar hookups, ranging between $0.002 to $0.007 per gallon, which means anyone taking water from the tap of this system should donate at least a penny for every 5-gallon bucket. Due to the lack of use of this system, the water probably costs us much more than that, perhaps over a penny a gallon, but to encourage self-sustainability, it is my feeling that we should charge less per gallon for this water than the city charges.

Update 2008-03-08:

I've done some testing of the solar well over the last two days. Draining about 30 gallons from it did not deactivate the float switch, so today I climbed up, pulled out the float switch, and dropped it back in. That made the float switch lamp go out, and the pump activated. It pumped for at least an hour, but I didn't hear any water going into the reservoir, and the float switch never activated. When I went back some time later, the pump would switch on for no more than a few milliseconds, then cut out again. This happened every minute.

So, I opened the top of the pump housing, which I shouldn't have done because I know nothing about how it's set up. There was a piece of yellow rope showing, which I pulled, then released, and it fell down the shaft, making a hissing sound as it went. That was the first "oh, shit" moment. Then I shut off the breaker, hooked up the winch, and started pulling out the pump. The winch is secured very flimsily to the pole, and before I have another "oh, shit" moment I'm going to see what can be done to secure the winch better. I hope to be able to leave this better than I found it, if not completely working, by the time I'm done. Your patience is appreciated, and if anyone can clue me in on what that piece of rope, and whatever is attached to it, is for, I'll appreciate that too.

Apparently I was pumping dry, which the control circuit isn't supposed to allow, and burned out the pump.

Update 2009-01-17:

Today Neal, Dave, and I got the solar pump out. The rope I dropped last fall turned out not to be a problem, it fell down past the pump and came out afterwards.

Dwayne Madsen has to say about the pump: "Rebuild it for a challenge, but I would not consider putting back in service. There may be no parts available, and still, it is a sub-par low output pump. The community would be better served with a new Grundfos and maybe a few more panels."

Nevertheless, I will try to rebuild it myself and put it back in service. If the community wants to spend money on a new pump, that's another issue.

Update 2009-07-26 and 2009-07-27:

I don't have any good news about the well, but I'm about to go on a crazy bicycle/windsurfing trip that I might not survive (though I definitely PLAN to), and it's only fair to let everybody know the status of the solar well.

The flex pipe and electrical wire haven't noticeably deteriorated over the past months they've been lying on the ground, but it'll all need to be sterilized before putting back into the casing, so when going over it with a cloroxed rag, it should be carefully inspected for damage. The electrical tape holding the wire to the pipe should probably be replaced in any case.

The pump is shot. I hastened its demise for sure, but it historically has needed service every few years anyway, and since I haven't been able to take it apart, let alone fix it, and since none of my get-rich-quick schemes have worked, I can't buy a new one. So if the community wants a source of water after TEOTWAWKI (the end of the world as we know it), it would be prudent to invest a few thousand on a new Grundfos or other good quality deep-well solar pump. Here is one that should work:

It has dry-run protection, and runs on just about any voltage over 30VDC. The solar panels as installed provide 48V to 60VDC on sunny days. As I mentioned before, the float switch inside the storage tank seems to be buggered as well, and will probably need replacing, as will the controller. And if you look at what happened to the bronze pipe fittings after sitting underwater a few years, you might also consider it wise to get stainless fittings next time.

Even this great pump probably won't last unless it's used. And it probably WON'T be used as long as grid power is providing PRESSURIZED irrigation water via the other well (this storage tank is just gravity feed). So unless people can religiously go and get a few gallons a day from it, it could atrophy just sitting there, as this one did. Then again, the "conspiracy nuts" are saying that all hell is going to break loose after the end of this September. So if we wait until after that, all bets are off for being able to buy and install the pump.

I'm leaving it up to the community to decide. My preference would be to have it fixed soon, IF people can haul water from it regularly. Of course it will need to be distilled before any internal use.

I forgot some things: since the pump was running dry, it's probably safe to assume that the well needs to be dug deeper, or the well has silted up, either of which will probably mean the services of a well driller are needed. Here are the specs given to me by Joyce about a year ago:

Total depth: 300 feet
Depth to water as of Sept. 4, 1994: 230 feet
Perforations from: 227 feet+
Casing bottom: 289 feet
Casing diameter: 6 5/8 inches