Version 1.0
Our well taps into the Sonoran Aquifer. Building on a hill limits the possible locations for a well, in addition to legal constraints imposed by Planning and Zoning. Finding the best place to drill involved a compromise between probable depth to the aquifer and distance from our storage tank.

The lowest point on our property is 200 ft below the highest, and, at about $15/ft for total drilling costs, looked pretty good … unfortunately, this point was also about 750 feet from our storage tank. The optimal location turned out to be on the north side the hill, about 100 feet from the tank. We got lucky and hit 70 gpm at a depth of 270 feet. That’s a lot of water. Still, the location was 100 feet from a power supply. We decided to put in a solar powered well pump.

Four 50W Siemens solar panels provide power for a submersible 48VDC piston-type pump made by Dankoff Solar Products, Inc. of Santa Fe. Of course, we can only pump for about 3-5 hours each day, depending on season, when the sunlight is strong. Still, the system performs well (no pun intended) with the pump delivering around 1 gpm to the surface. This is more than enough to meet our yearly average water demand of 20 gpd. The 2000 gallon storage tank provides an ample safety margin for extended cloudy periods. Cost for the pump was $1800. Cost for the solar panels, rack and slab was about $1200.

Version 2.0
Well … we thought that was the end of the story, but there’s more to relate. Of course we monitored the water yield closely over the first year of operation, even though the system was designed to meet our average needs and could theoretically be “out of sight and out of mind.” Properly designed solar energy systems should require little maintenance or attention. As it turns out, during our peak demand season (summer) when we should be getting plenty of solar energy, we fell short of meeting our water demand.

Here’s the problem. When a system like this is designed, one of the key design factors is a variable called insolation, a fancy term for how much sunlight to expect on an average day. One obtains this number from a geographical chart based on meteorological records. For Phoenix this number is given as 4.8 kW-hr/m²/day (designing for worst case scenario). As it turns out, this number is no longer valid. Being an amateur astronomer, I should have known better. When we first moved to Phoenix in 1978 this number probably was valid. On average, 1 out of 7 planned astronomy events was cancelled due to cloudiness (15%). Over the last five years that number has changed to more like 1 out of 5. Indeed, it was during a protracted two-week cloudy spell in July of 2001 that our water reserve fell to under 200 gallons. Why the decrease in available sunlight? Answer: climate change, both global and local. There’s the proof for you in hard numbers. And it cost us hard cash to redesign.

We added four more 18W panels to increase our pumping power. This cost an additional $600 over what we’d have paid for four 75W panels had we done it right the first time. Now we have all the water we need. The moral of the story is: research your local weather data over the last five years and adjust the insolation values accordingly. In the desert southwest, which is getting more humid, that reduction is about 25%.

Version 3.0
Déjà vu. There’s still more to tell. Our first pump failed in 2004, requiring replacement. At that point, new pump technology had become available in the form of a “brushless” pump design with a claimed longer service life. That was hard to resist, since pulling up 300 feet of heavy pipe to service a pump is no small task. The new pump, an ETAPump by Lorentz, also offered a higher efficiency electrical power system that could deliver water at a faster rate. Unfortunately, it also required more power. So we upgraded the PV array one more time to two Sharp 165W panels. The system has delivered water flawlessly since that time (knock on silicon). If you’re wondering what became of the old solar panels, look here and scroll down to “Let There Be Light.”
Version 4.0
It was with a heavy heart that, in late 2009, we had to replace our solar pumping system with one powered by standard AC electricity. After nearly 10 years of pumping our own water with free energy from the Sun, we noticed that the water level in our 2000 gallon storage tank was starting to drop. Previously, we were able to pump enough water to keep up with our demand. We don’t use a lot of water, having low water use appliances, xeriscaping instead of a lawn, and no swamp cooler … only about 30 gallons/day are required. Unfortunately, between the continuing drought conditions in the Southwest, and about a dozen new wells drilled up-aquifer of us, the water level in the aquifer was dropping.

When we first drilled the well, the static water level was at a depth of 250 feet. It is now at 264 feet — a drop of 14 feet in the aquifer over 10 years. The recharge rate has also dropped from its original 70 gallons/minute to something around 0.1 gallon/minute. That’s still enough to meet our demand. But here’s the real problem: when you pump with solar energy, you can only pump for about 6 hours/day, and even then, only if it’s not cloudy. My neighbors are all pumping 24/7 using AC powered pumps, so they are “stealing” our fair share of whatever water remains in the aquifer. Thus, we made the decision to switch over to AC pumping. It required extending an electrical circuit from the house out to the well, replacing the solar pump with a conventional AC pump, and a “step backwards” philosophically.

But here’s some good news. Arizona Public Service, our electricity provider, offers “time of use” rate plans. Electricity purchased during peak demand hours (9 am – 9 pm) costs us about 4 times as much as that purchased during off-peak demand hours (9 pm – 9 am). So we added a timer module to the well pump circuit, such that it only runs during off-peak hours. That means we’re only pumping 12/7, but we can adjust the timing as needed to respond to the aquifer, and still reduce our energy costs for water.

As of October 2009, we are pumping enough water 24/7 to keep up with demand, and our storage tank remains topped off, as it always used to be. The switchover to AC wasn’t cheap, but water delivery via tank truck cost us $60 to fill our storage tank, and you know that’s only going to get more expensive over time. [As of 2023 the cost is $120.] When you live outside the city you have to find your own water source. Obviously, rainwater collection systems are another option, and they work great in the Pacific Northwest, but not so much here in the Sonoran Desert.

[Update 2014] Aquifer production continues to drop. In an effort to keep supply up to demand, we removed the timer from the circuit and are now pumping 24/7. To reduce wear and tear on the pump, we set it to cycle on every 4 hours, by which time enough water has accumulated in the well to make it worth pumping.

So what you see in that last photo is all that remains of our once beautiful solar pumping station. Just some electrical conduit and a new set of controllers. We sold the Sharp 165W panels locally (at a loss) — PV prices continue to drop, and those panels were only 3 years old. It’s a sad ending to this story, but water is priority #1. You get it however you can.