New China Shop Topic: How Accurate is the Pool Calculator?

[CarlD]

Today I saw this post from one of our regular posters, AnnaK, whose contributions are always welcome.

Those test kits are fun, aren't they?

At the bottom of the PoolCalc there's a feature called "Effects of adding chemicals" .

It says that adding 180 oz of 6% bleach (the 2 bottles you bought) will raise the FC in a 36,000 gal pool by 2.4 ppm.

Looks like you have things well under control and are on your way to a 5 Minutes A Day pool.

—AnnaK—

So I decided to manually calculate the effect of adding bleach in this case. I was very surprised. But rather than starting a debate in a thread helping a pool owner, I figured the China Shop was the place for it.

Just to point out at the start: Accuracy and Precision are not the same thing. In fact, they are opposite concepts. Accuracy is a description of how far off you can be, where as precision is a measure of how close you are.

Hate to tell you this but I just found that the the Bleach Calculator is inaccurate (I hesitate to say "wrong"--which is not the same thing) due to rounding off significant digits.
The correct increase is 2.3ppm, not 2.4 (actually 2.34ppm)

Pool: 36,000 gallons
Bleach strength: 6% or .06
Amount of bleach: 180 ounces or 1.406 gallons.

The underlying formula for calculating FC delivered is:
(1,000,000/36,000)*.06*1.406 = 2.34ppm, where .06 is, of course the bleach strength, 1.406 is the amount of bleach.

But the calculator rounded the bleach amount up to 1.41 gallons, which results in a ppm increase of 2.35, which it then rounded to 2.4.

So by rounding the amount of bleach up, it pushed the PPM to the rounding threshold, and then rounded up again.

Does it matter in the big picture, ie, your pool? No.
It's probably not actually 36,000 gallons either, since a drop of 2" in a 20'*40' pool is 1000 gallons, which would change the final measurement, too, and make it 2.41ppm.

But in all things B-B-B we take nothing for granted or as gospel.

Now I realize that we all want things to be easy and fast in pool care. But there's a big difference between knowing we are ball-parking a number or amount, and assuming a calculation is correct when it is not.

I ball-park all the time. I use "20,000" gallons in 99.99% of my pool calculations, when it can vary from 18,500 gallons up to 19,600--just 2-4 inches or water can make that big a change. Plus the 19,600 is just an estimate as well.

I also use my "Rule of Thumb" calculation of bleach usage and ballpark around that. But again, I KNOW I'm deliberately allowing a reduced level of accuracy. Still, the basic assumption of the rule of thumb introduces, at its root NO inaccuracy.

It assumes that a pool of 10,000 gallons IS 10,000 gallons, to within a gallon (ie, 9999.5 to 10,000.5), that the bleach concentration is to within ten-thousandth (.05245-.05255 for 5.25% bleach), and that 1 gallon is one gallon, to within a half-ounce either way, and therefore that the values of 10,000, .0525(or .06) and 1 can be used with confidence.

Clearly though, the calculator is making rounding assumptions that introduce an unnecessary (IMHO) decrease in accuracy.

Does this mean we should throw the baby out with the bath water? Definitely not!
But it does mean that if it is critical to have extremely precise results, you should consider using the formula manually.

Carl

[AnnaK]

The short answer to the question then is YES. It's accurate. It doesn't claim to be precise.

Our test results go to 1 significant figure on the FC/CC test. Sample sizes likely vary from one time to the next by a ml or some fraction thereof. Pool volume is certainly not static. Bleach concentrations differ between a fresh bottle and one that's partially used. Dropper sizes change depending on humidity and static electricity. An operator's visual acuity changes with ambient light. And most importantly, goals and targets for a "balanced" pool are given in ranges, not in numbers to 4 places beyond the decimal.

'Fess up: do you REALLY care whether you raise your FC by 2.3 or by 2.4 ppm?

[CarlD]

"Fess up.."

OK. Full Disclosure: I confess. I really do NOT care if it's 2.3 or 2.4!

Saying that, I'm well aware of all the other variations. In fact, I spelled them out. I know how to test for chlorine concentration (I've posted the method a number of times). I know how to calculate pool volume (most use 7.5 gallons for gallons/cubic foot but the more precise number is 7.48). I even realize that drop size changes and the recommended way is to hold the bottle exactly perpendicular.

But I have shown how two separate rounding functions used consecutively have generated an inaccuracy that results in .1 increase above the correct rounded figure.

Is accuracy to 4 decimal places necessary? Sounds petty to say "yes" but it's not. Bleach concentrations go into the formula out to 4 decimal places routinely (5.25% bleach is .0525 in the formula).

We estimate pool volume. We estimate bleach concentration. We even estimate amount of bleach. These are all, to the statistician, known levels of error or "noise" The formula should NOT introduce another level of "noise" if it is not necessary.

Your statement at the top "YES. It's accurate." needs to be amended: Yes, it's accurate to 1 part per million. This is certainly true and I have no problem with that.

However, the calculations CLEARLY show that it is NOT accurate to 0.1 part per million.

Carl

[AnnaK]

Which of the water balance parameters require 0.1 ppm?

Or, coming at it from another way, do the many variables we've both enumerated need an accuracy consideration to 0.0001 part per billion? Can we even measure that?

[CarlD]

Which of the water balance parameters require 0.1 ppm?

Or, coming at it from another way, do the many variables we've both enumerated need an accuracy consideration to 0.0001 part per billion? Can we even measure that?

Don't you mean 1 part per billion (1/1000 of a million).

Of course none of the majors do (pH doesn't really count). The lowest we EVER need to go is .2 (useful if CC is not clearing properly).

But you are defending a known, fixable inaccuracy in the system.

Carl

[Manslick]

What if 6% chlorine is 5.9%. OMG.

[AnnaK]

The inaccuracy of the bleach calculations within the Pool Calculator is not really germane within the context of usage.

The defense rests. The defense is just about to take her dogs swimming in a pool that may well be off by a couple parts per gazillion (that's 10 to the nth). The defense does not plan to agonize over that. Instead, she plans to dose herself with an imprecise amount of Margarita, accuracy be gosh danged.

Have a great weekend!