I know there is a long list of "interferences" for ORP measurements including CYA, pH, etc.

It appears that there is a complex relationship with temperature as well. I get very large swings in ORP readings as my sample of pool water warms up from ~60 degrees to ~75 degrees. I say complex because it appears to be very nonlinear.

Is there any data on the temperature dependence of ORP for the temperature ranges and chemistries common in pool water?

I'm using a Hanna pH/ORP combo probe. The pH readings are great. I keep the probe wet with pool water and the calibration has been very stable as checked periodically with reference solutions.

The ORP readings require warming the pool water while the water is so cold. This shouldn't be an issue over the summer. With the water warmed, I get pretty consistent correlation between FC and ORP.

Looking forward to swim season!

Marc

ORP readings are dependent on the electrical potential of the water, which changes with temperature. Here are some google results that do a bit of 'splainin on just what that effect is.

The theoretical Nerst equation for the relationship of electrochemical potential vs. temperature and concentration is as follows:

E = E₀ - (RT/nF)*ln(Q)

where Q is a ratio of product to reactant concentrations. R is the universal gas constant 8.314 472(15) J/(K mol), F is Faraday's constant 9.648 533 99(24)×10⁴ C/mol and T is the temperature in Kelvin which is the temperature in Celsius plus 273.15.

In practice, the ORP reactions don't seem to exactly follow the Nernst equation because the actual half-reactions apparently aren't simple and clean. I've fit empirical formulas to Chemtrol sensor data and to Oakton sensor data as follows (these are in my spreadsheet):

ORP(mV) = 1308-1000*(LN(10)*8.314472*KelvinTemp/((2.6-0.24*pH)*96485.3415))*(-LOG([HOCl])+(2.6/1.25-0.24*pH)*pH)

ORP(mV) = 1399-1000*(LN(10)*8.314472*KelvinTemp/(0.635*96485.3415))*(-LOG([HOCl])+0.24*pH)

In practice, this results in the following ORP for each sensor vs. temperature at 3.5 ppm FC with 30 ppm CYA and pH 7.5

TEMP ... ORP (Chemtrol) ... ORP (Oakton)
.. 40 .......... 740 .................. 710
.. 50 .......... 729 .................. 696
.. 60 .......... 718 .................. 682
.. 70 .......... 706 .................. 668
.. 80 .......... 695 .................. 654
.. 85 .......... 689 .................. 647
.. 90 .......... 683 .................. 641
.. 95 .......... 677 .................. 634
.. 100 ........ 672 .................. 627

So roughly speaking, the ORP for a given sensor seems approximately linear in the narrow temperature range of pools and spas, but the actual ORP mV difference per temperature difference varies by sensor. However, it sounds like you are seeing something different, probably because you are starting with much colder water.

There is a temperature dependent formula for chlorine with cyanuric acid and if I account for that then the above table becomes the following:

TEMP ... ORP (Chemtrol) ... ORP (Oakton)
.. 40 .......... 654 .................. 602
.. 50 .......... 667 .................. 618
.. 60 .......... 679 .................. 634
.. 70 .......... 691 .................. 649
.. 80 .......... 701 .................. 663
.. 85 .......... 706 .................. 669
.. 90 .......... 710 .................. 674
.. 95 .......... 714 .................. 679
.. 100 ........ 717 .................. 683

So what actually happens seems very dependent on what's in the pool, such as CYA. Essentially, the relationship of chlorine and CYA is such that at higher temperatures the chlorine binds less strongly to CYA so the active chlorine (hypochlorous acid) concentration is higher at higher temperatures and this overwhelms the normal declining electrochemical potential as the temperature rises.

So that I can improve my modeling, what sort of ORP readings are you getting at different temperatures? Do you see the ORP fall or rise as the temperature gets higher (at the same FC)?

Richard

Thanks Richard,

That is interesting information. I'll collect some real data, but my sense is that I am getting more like a 100mv increase warming from 55-60 to 75-80. What I am now thinking is that in addition to the temperature rise, the other significant variable may be keeping the sample out of the sunlight. How long after removing UV exposure does the CYA release the chlorine? My CYA is probably around 40 ppm.

The release of chlorine from CYA is fairly fast. The half-life for one of the reactions is 1/4 second and for another it's 4 seconds. In practice, things equilibrate in the time of your testing. The ORP is measuring more of the instantaneous state of things though there is a very small current that occurs even when measuring voltage. Your FC test measures all the chlorine and the chlorine bound to CYA gets released well within the time of the test, though maybe at colder temperatures it slows down enough to just be noticeable (it would look as if the FC increased a bit over a short time using a DPD or OTO test, while with FAS-DPD it would be like a fading endpoint if you added drops quickly enough).

The UV from sunlight would decrease the chlorine in the water a very small amount relative to the speed that chlorine is released from CYA. Though people have measured ORP decreases waving a UV lamp over a cell, that's a lot more UV exposure than from the sun in a pool.

Richard

Such a short release time surprises me. I always understood that the UV binding of chlorine to CYA had a large impact on the availability of chlorine to sanitize, hence the higher FC required with higher CYA concentrations. But only a small volume of water is exposed to UV since it is absorbed by first few inches of water. With a short release time, the bulk of the water is really not affected by UV.

Its always good to reexamine what we think we know.

First of all, the binding of chlorine to CYA has nothing to do with UV. Where UV comes into play is that the chlorine bound to CYA as well as CYA itself tends to absorb UV without getting destroyed by it (i.e. it basically converts it into heat). Unbound chlorine is what is susceptible to UV where roughly half breaks down every 35 minutes in direct noontime sun.

Also, UV does penetrate into water deeply -- it is infrared that gets absorbed a lot by the water in the first few inches. When there is CYA in the water, however, it does absorb UV though I don't have data on how much for sunlight so don't know how much it penetrates into depths. Since higher CYA levels seem to protect chlorine from breakdown even keeping the FC/CYA ratio constant, I speculate that CYA does shield lower depths from UV, but I don't know if this is really true nor do I know the size of this effect.

As for chlorine getting released relatively quickly from being bound to CYA, you need to distinguish between the killing power of chlorine which is based on reaction rates which are based on the instantaneous concentration of active chlorine (hypochlorous acid) vs. the chlorine that is bound to CYA which needs to be seen as a "reserve" of chlorine that is inactive. The fact that it gets released quickly is irrelevant to chlorine's killing power. Think of an army with only the front line having rifles and able to kill the enemy. It doesn't matter how quickly you replace a fallen soldier (one that has sacrificed himself killing the enemy) with another one that comes from the reserve -- the rate of killing is solely a function of how many active front-line soldiers there are and not how many are in reserve nor how quickly ones in reserve replace those on the front line. The amount in reserve, which is mostly what is measured by FC, is important to know that you have enough chlorine to not "run out" as it gets used up, but by itself is a meaningless number with respect to chlorine's killing power (i.e. you need to know CYA and pH to be able to calculate its killing power).

The fact that ORP is so significantly affected by having CYA in the water is because it is mostly measuring something proportional (in a logarithmic sense) to the "active" chlorine (hypochlorous acid) concentration. It really isn't measuring FC. At the same FC, but with no CYA in the water, the ORP reading would be vastly higher -- in fact, too high in that the chlorine would be too reactive and would noticeably affect swimsuits, skin and hair over time.

Richard

ORP readings are sensitive to sunlight in some way, though I have no idea why. ORP readings go down by about 30 on my sensor when the sun is shining on the pool and go back up when the sun goes behind the trees in the late afternoon or when it gets very cloudy. You can duplicate the same effect with a portable sensor and a cup of pool water by taking the sample in and out of direct sunlight. Even small amounts of indirect sunlight can have a smaller but still measurable effect.

My very casual experiments show that the reading goes down when exposed to sunlight fairly quickly, less than ten minutes, but appear to go back up more slowly, perhaps 30 minutes or an hour after removing from sunlight before it is back to what it started at. This effect appears to depend on CYA, lower CYA levels show a less obvious effect.

I don't think this effect is because active chlorine is getting destroyed by sunlight and then replaced more slowly by chlorine getting unbound from CYA. The time constants appear to be wrong for the observed effect, though I never corrected for indirect sunlight effects, nor did I use a fine enough time scale to be completely sure of my results.

ORP readings also vary because of several other effects. Dissolved hydrogen gas from a SWG can lower ORP readings. The rate of outgassing is fairly slow, several hours for a cup of water, sometimes much longer for an entire pool. The hydrogen gas effect is quite large in my pool, which seems to be somewhat unusual, drowning out the FC signal. Temperature and PH also affect the reading. Both of those can change enough to have a noticeable effect while a sample in a small container is sitting out.

The lesson I take home from all this is that ORP readings and active chlorine levels are not linked anywhere near as tightly as some ORP vendors would like you to believe. ORP reacts to a number of factors that have nothing to do with the effective sanitizing ability of the chlorine. Ideally all of these other factors are held constant, so that the FC level is the only thing affecting the ORP reading. In the real world it is often, but hardly always, possible to get the other effects to be held close enough to constant for ORP to be useful.