I've previously talked about the kind of collection efficiencies that rain barrel users can expect. They are typically dismal unless you have a very large barrel or the right kind of rainfall pattern. Collection efficiency depends on the rainfall pattern since one large storm would overwhelm most storage systems, thereby wasting the majority of the rainfall whereas a large number of small storms might never once overflow a typical barrel system. In the latter case efficiency would be 100%.
Based on typical southern California rainfall patterns I estimate a typical rain barrel collection efficiency at 10-20%. This is based on my previous analysis after throwing out the highest and lowest rainfall years and applying a factor of two reduction in efficiency, since the numbers I calculated were best case, and nobody is that diligent. Susan Carpenter of the LA Times spent $500 on two rain barrels and received a third for free. If the roof area she captured rain from was 1500 sq ft, then her cost per square foot would be $0.33 for 10-20% efficiency. She was happy to pay it, and I guess that speaks to the psychological impact of having a seemingly large amount of water deposited, as if by magic, in your barrel.
But is 100% efficiency achievable and cost effective? Yes, just about.
We know that some city codes require new construction to keep all* rain water on site. In fact, Bad mom recently blogged about the new Redondo Beach Library that has a swale and infiltration area intended to do just that. Let's call the collection efficiency of this engineered infiltration area 95% to account for the fact that "all" is probably not really all. Here's a Google Maps snapshot of the Redondo Beach Library being built. The infiltration area will end up being at the south end of the dirt lot and will infiltrate rain from 25,000 sq ft! You can estimate the size of the lot too, based on the 50x100' lot sizes of the surrounding homes.
If Susan Carpenter's roof area is 1500 sq ft and that required 3 rain barrels to get 10-20% efficiency, how many rain barrels are needed to capture runoff from 25,000 sq feet of lot? At 95% efficiency? The answer is far too many.
Suppose the infiltration pit at the North Redondo Beach Library added $20,000 to the cost of the project (this link says the whole project was $6 million so this is 1/3 of 1%). I have no idea what the actual cost increase was, but this seems like a good starting point for a construction project of significance. The lot is 25,000 sq. ft. as I noted above and they capture 95% of the rain from that lot. Then the cost per square foot is $0.80. Don't forget: That's at 95% efficiency! If we all agree that we ought to save rainwater, then shouldn't we be happy to pay $0.80 per square foot when the efficiency is so much higher and the area served is so much greater than that achievable with rain barrels?
The cost numbers get more lopsided if you factor in efficiency directly. Then we have a cost of $2.22 per square foot per point of efficiency for rain barrels and a cost of $0.84 per square foot per point of efficiency for infiltration.
You say to me, "I'm just a homeowner. I don't have $20k for an infiltration pit. All I have is $500 for some measly rain barrels" And I'd have to agree, $500 isn't too much. But it's not nothing. Unskilled labor costs $10-15 per hour, so with $500 you and your unskilled labor could regrade your lawn to include a slight berm on the downhill side and direct your roof runoff onto the lawn. The details are probably long enough for another blog post, but it's money better spent than on rain barrels.
And there's no magic involved.
*"all the rainfall" of course has an engineering definition that probably has to do with the mean rainfall in a storm plus some margin so in a very heavy storm not all the actual rain runoff will be captured.