Trying to keep up w/ Chem Geek

[chem geek]

Take a look at this PDF file and look at the EFFICIENCY vs. FLOW chart. Though one can debate whether this is accurate, it's at least a starting point for discussion. According to this chart, the solar panels are 80% efficient at the recommended flow rate of 4 GPM per panel. The maximum flow rate of the panel is 8 GPM and the minimum is 3 GPM. The max and min are recommendations. The panels themselves can handle an awful lot of pressure so the maximum can probably be exceeded without a problem, but there might be issues with the "header" that distributes water evenly if the flow rate is too low, BUT I DON'T REALLY KNOW.

At 2 GPM the efficiency is about 70% while at 1 GPM it's about 60%. So let's do an analysis at 2 GPM as an example, comparing against 4 GPM. My own system is rather large with 12 panels so at the normal 4+ GPM that is 48+ GPM total (my panels are connected in parallel) while 2 GPM per panel would be 24 GPM total. For the Inteliflow (and 4x160) pump, I determined the pump curves to be determined with a pretty good fit by the following equation:

Head (in feet) = (RPM/350)^2 - (GPM^2)/470

and the output power is given by

Output Power (in watts) = Head (in feet) * RPM * 0.188165

Near the peak efficiency point, the pump is about 50% efficient and this point occurs roughly where the specific speed of the pump is at 1320 per the following formula:

Specific Speed = RPM * sqrt(GPM) / (Head^0.75)

Now things get tricky to calculate since we need to determine the system curve. The panels don't add much to the Head and have the following head loss formula (remember that the panels are connected in parallel so the loss for one panel is the loss for the system of panels):

Solar Panel Head Loss (in feet) = (GPM^2) / 8

The cartridge filter I have has a head loss curve as follows:

Cartridge Head Loss (in feet) = (GPM / 62)^2

Though the actual head loss calculation for a 2" (nominal) pipe (the size of all of my piping except for the suction side) is complex, I can see that it is approximated by the following formula:

2" Pipe Head Loss (in feet per 100 feet) = (GPM^1.8) / 295

and the suction side with TWO 1.5" pipes is approximated by the following formula (for each pipe):

1.5" Pipe Head Loss (in feet per 100 feet) = (GPM^1.8) / 90

so the actual GPM to use for the 1.5" pipes is half of the system GPM since there are two suction pipes right up to the pump.

Let's assume 100 feet for each 1.5" suction pipe and 400 feet for the 2" pipe on the output side of the pump. Then our expected head loss at 48 GPM and at 24 GPM will be as follows:

48 GPM: 4*(48^1.8)/295 + ((48/2)^1.8)/90 + (48/62)^2 + ((48/12)^2)/8 = 21.7 feet

24 GPM: 4*(24^1.8)/295 + ((24/2)^1.8)/90 + (24/62)^2 + ((24/12)^2)/8 = 5.8 feet

Using the Inteliflow equation and solving for RPM (which we don't really need, but I want to cross check against the Intelliflo curves) I get:

RPM = 350 * sqrt(Head + (GPM^2)/470)

48 GPM: RPM = 350*sqrt(21.7 + (48^2)/470) = 1805
24 GPM: RPM = 350*sqrt(5.8 + (24^2)/470) = 928

The output power is:

48 GPM: Output Power = 21.7 * 48 * 0.188165 = 196 Watts
24 GPM: Output Power = 5.8 * 24 * 0.188165 = 26 Watts

Assuming that we are roughly hitting the 50% efficiency spot on our pump power curves, then this means that whereas before we were at around 400 Watts, with half the GPM rate we are now at around 55 Watts. However, there is an overhead of power lost in the pump windings and this appears to be around 80 Watts so in reality our output power is probably something closer to comparing 450 Watts vs. 130 Watts or a savings of around 70%. We only dropped in efficiency of the solar panel from 80% to 70% so clearly this is a large savings overall in electricity costs even if you run the pump 14% longer to make up for the loss in solar panel efficiency.

Note that to make all of the above work out well, you have to have a variable speed pump. You COULD try just using slower flows and downsizing a single speed pump, but it's harder to hit the sweet spot. With my current pump and piping situation, I actually see nearly 30 PSI at my filter gauge so including suction loss it's probably 32 PSI or so actual total head loss (that's about 75 feet!) with a 65 GPM flow rate (based on my current pump curves). This implies an effective output pipe length of closer to 1000 feet so something is very very strange about my system that I haven't figured out yet (perhaps all those twists and turns in the pipes are adding up to much more than I think).

Note also that cutting the GPM rate in half means that it takes longer to have a full turnover of your pool. In reality you need to run your pump twice as long, not just 14% longer (to make up for the solar efficiency). So when dropping the GPM rate in half from 48 GPM to 24 GPM you are really comparing 450 Watts against 260 Watts for equivalent run times so the savings is really around 40%. Due to the fixed power consumption "winding losses" (and other losses) of the pump regardless of RPM, it doesn't make much sense to go much slower.

Richard