Solar pool pumping makes even more sense


8 years ago, I proposed that pool pumps really should be designed to make use of solar (or wind) power. We have now seen used solar panels show up at the ridiculous price of $50 for 250 peak watts.

What this means is a solar powered pool pump could do most of the pumping with solar for what it costs for just a few months of electricity. These are not high quality panels. They are so cheap that it would reasonable to hand them on the north fence of the pool area even though vertical hanging is less efficient, though the roof would be great. (Hanging on a south facing fence means getting only 60-70% of what you get from optimal tilt on a roof, but at this cost it could be a win, though sadly performance is worst in summer.)

(Panels could actually be designed for vertical mounting, with the cells mounted on a sort of staircase platform which tilts the cells close to latitude while making the hanging structure just a few inches thick. One could make an adjustable angle frame which eager people could actually adjust a few times a year for even better efficiency.)

With solar this cheap, and getting cheaper, we face the problem that the other costs of solar (mounting, labour, wiring, permits, inverters, grid-tie) now dwarf the cost of the panels. At 20 cents per peak watt, the other costs are now an order of magnitude greater. The big opportunity comes in finding ways to eliminate these costs with easy self-install or basic contractor install.

This means you want loads you can power without grid-tie. As soon as you grid-tie you immediately need a whole level of special gear and permits. You are playing with dangerous stuff.

You also want loads that can deal with intermittent power. Pool filtering is ideal for that. You don't care when the water is filtered. You don't care if the filter stops when a cloud goes over. You just care that you get enough filtering to keep the water clean, on average. It's OK to filter more one day and less the next. There are few loads that are large and meet that criterion, particularly in the home. Charging EVs would meet the criterion except you want to disconnect the EV often during the day to drive it places.

All you need is a pool pump that can take the power directly from the panels by DC, and filters at whatever the sunlight can provide. When it's sunny it filters at a good speed. When it's night or raining it doesn't. It also turns out that you need more filtering in summer and less in winter, which corresponds well to how solar panels provide output, though less so for vertical ones on a fence.

The pump would also be able to switch over to the mains, ideally to add to the solar, but more simply just temporarily disconnecting them. You would use that when running a pool vacuum that needs high flow or other special operations. You would also do that if the system detected the sun just didn't provide enough filtering in recent days (and when forecasts say the sun isn't coming tomorrow.)

Because it would just switch power, or combine them at the DC level, the system would never feed power back to the grid. It would be unable to. Low-level filtering only needs about 350 watts typically. The electrical wiring would be simple and safe, on the order of a basic extension cord.

Panels could be nailed on a roof, hung on a fence, or just put out on spare land for those with large lots, in that case tilted to latitude.

Pool owners are often spending $30-$60/month in electricity for their pool pump. The panels and pump modifications (which should not add very much to the cost of a pump) would pay for themselves in months, not years. Even an add-on box on the power circuit to a modern smart pump could do the job.

The closest thing you can get to this are floating pool filters with solar panels. These aren't very popular, because they need to be big, and even then, can only do about 1/4 of the filtering job. It's a nice idea, self-contained, but ideally you want 20 square feet of panels, which is too much to float in a pool. Also you can't have a pool cover as easily.

(As I noted in the prior article, wind is also a perfect match for pool filtering -- no surprise since using wind to move water is ancient stuff -- but wind prices have not fallen through the floor like solar has.)

A new era of solar -- too cheap to be optimal

We now enter an interesting era of solar, where panels are so cheap that it is wrong not to waste them. What matters with solar is the cost of the whole system over a period divided by the kwh the system will deliver in that period. So even if vertical panels only deliver half the electricity of properly roof mounted ones, that's still a green win if you save more money not going on the roof than it costs to buy extra panels. If you have to throw away 2/3rds of your power to avoid grid tie, but it's 80% cheaper to not grid tie, it's a win.

That's because what matters is how many kwh you unload from the dirty grid per dollar you spend. That's the way to be green, to remove the most fossil from the grid with the money you have.

That may seem at odds with what I have written about solar panels on cars. They are wasteful (horizontal, not always in sun, useless when battery is full) but they are also expensive (need robust environment.) On the other hand, they don't need grid tie. But if you already have grid tie, then more panels on the roof or in a solar farm do much better at kwh taken off the grid per dollar.

Other loads that can use solar

Another load that could use solar is any constant load, like computer servers. With an always on device, you just need a power supply that takes any solar power it gets and supplements any shortage from the grid. When the sun is high, the device will be running entirely on the solar. At night, it will run entirely from the grid. If the panels match the load at peak, you use 100% of their power. Again, no grid tie, there is never surplus. This could also be done for all the always-on parasitic loads of a house, but it's hard to reach them all. Most of the loads in a house are intermittent, and you care about when they are on or off. To use those with solar you need either grid tie or batteries, and these are now the most expensive part of the system.


The tech for making a hub that switches loads to mains when solar is not supplying enough power is cheap and easy and avoids so many problems (mostly regulation) of grid tie systems. Sadly I still have not seen a well designed and priced system on the market, although there have been many fake systems sold (or at least paid for).

Yes, I have also been hoping for this. One way to do it that I describe here is in the standard PC power supply. Just make one that can take solar panel inputs (and even connect a battery.) PC power supplies already have the DC-DC and AC electric power conversion tools and smarts. When I wrote about this it was more common to have a PC that was always on, but home devices are getting better at falling asleep.

That's the problem with solar in a box. The devices powered by it need to use all the solar in order to be efficient. Which means they need to draw a constant load that matches the peak load of the panels during the day. If not, you are just throwing away a lot of your solar power.

The problem is there aren't many loads like this. Pool filtering is one of them, because it's the one load that doesn't care when it gets power or how much. But an always on PC is also drawing 100w all day, as is an always on TV or monitor. All the parasites in the house are probably drawing 100w in some houses -- but they are plugged in all over the house.

Now "grid tie plug in the wall" may require a dedicated breaker, but more modern homes often have that. Another alternative would be to use it on a non-dedicated circuit but to put current measuring devices on all the other plugs on that circuit which let the device understand all loads, and never feed too much power into the circuit. The key, as you know, is plug and play.

What changed that made me write the article above is panels are now getting cheap enough that it is the right thing to do to waste panel. That was never true before. What matters is total installed dollars and effort per delivered kwh. If you have to buy twice the panels (because you hang them on a fence) but it cuts your install cost to zero, it's a win now. Now it's even OK to have your panels feed an intermittent load, wasting all their power when the load is off, if the panels are cheap enough and the duty cycle is high enough.

One imagines one could try to design an air conditioner to deliberately be run at 80-90% duty cycle just to run it on direct non-grid-tie solar supplemented with grid. I don't know if such a duty cycle is practical. Even so the panels are wasted all winter so it's not great.

The pool pump is the clear winner, and all it takes is for a pool pump manufacturer to make such a pump, or possibly to make a solar box designed to support today's smart pumps, which have an interface that lets you do remote control of their speed. I have such a pump and would definitely buy that. My efficient pump is now perhaps only costing me $25/month in electricity but even then it doesn't take long to pay for panels for it.

An A/C in the summer could act as a heat pump in the cold down to 30 or 40 degrees and a resistive heater below that. Probably won't be a replacement for burning fossil fuels in the coldest places, but it can at least cut it down so the energy isn't completely going to waste.

But now you're adding more expense and maybe the money is better spent elsewhere.

Add new comment