[az2008]

5 years ago I was trying to understand dissolved eggshells & how much Ca they create when mixed into water (for use in horticulture, and how it would contribute to "PPMs").

That got me over a big hurdle understanding fertilizers. Since then, I've realized "PPMs" (as measured by pen meters) are just an idealized notion of the conductivity of water. I.e., "if you dissolved 0.5 grams Sodium Chloride in 1L pure water (at a particular temperature), it would have *this* conductivity. So, *this* conductivity (whatever's in the water) is 500 ppm (displayed on the LCD screen of the $10 "PPM" meter) -- regardless of what's actually in the water.

My new question: What if I know (from the fertilizer packaging's "% of weight" analysis) that I'm creating 200ppm of N, 100ppm P, etc. Shouldn't I be able to calculate the expected conductivity of the water?

Relatively inexpensive TDS/EC meters exist which display the microsiemen value of the water (no conversion to idealized PPMs). Shouldn't I be able to calculate the expected microsiemens for a solution containing a variety of elements (like N, P, K, Mg, Ca, S, etc.)? Aren't there absolute values that I could use to extrapolate 200ppm N into an expected usiemens?

Basically I'm coming at it the opposite way most people would. They'd buy a $10USD PPM pen, calibrate it to 342 (or 1000ppm) NaCl solution. They'd put 1tsp of fertilizer into 1L water and read 700ppm on the meter. But, there's all kinds of abstractions happening in that scenario which make it more like a fairy tale than reality. Volume of fertilizer is meaningless (relative to the "% of weight" info printed on the bag). Additionally: all of the resulting PPMs come from everything except NaCl.

If I can use the fertilizer label to calculate expected PPMs of each mineral/element, why can't I predict the total microsiemens expected from that? Then use a meter that displays usiemens (bypassing the idealized representation of ppms. I mean, I know the PPMs I'm adding (based upon the weight of the fertilizer & its "% of weight" guaranteed analysis). I don't need a meter to tell me that. It would be more helpful/accurate/confirming if I compared expected conductivity to actual).

I hope that makes sense. I just need to be pointed in the right direction again. I need a table of "N has this conductivity; P has this conductivity." Something I could use to convert calculated PPMs (from the fertilizer label) to an expected conductivity. (What makes NaCl or KCl special? If those are known to a have a certain conductivity, why not everything else? There's a proprietary "442" scale which is made using various percentages of 40% sodium bicarbonate, 40% sodium sulfate & 20% chloride. If they can do that, why can't I calculate the expected conductivity of a more complex mixture of elements?).

[Borek]

My bet is that the only difference between meters reporting in ppm and μS is the scaling and the sign printed on the display (ppm or μS). Other than that they are identical.

This is in general tricky. If you assume you know the ratio of all ions present, conductivity tells you what is their total concentration. If the ratio is unknown, same conductivity can be a result of infinitely many combinations of the ions. One measurement is simply not enough to tell you anything.

Think about it this way: I tell you I have $100. You know the sum, but you have no idea if I have one $100 note, or ten $10, or $50, two $10 and the rest in nickels, dimes and quarters. Same about the composition of the solution that is 500 ppm (or whatever the equivalent μS is).

[az2008]

If you assume you know the ratio of all ions present, conductivity tells you what is their total concentration. If the ratio is unknown, same conductivity can be a result of infinitely many combinations of the ions. One measurement is simply not enough to tell you anything.

Thank you for replying. Regarding the bolded part: I think I do know that ratio (within reason?):

• I start with reverse osmosis filtered water (10-20ppm. I don't know what those ppm are.).
• Fertilizer products have a "guaranteed analysis" as "% of weight."
  
  • If I add 1 gram of fertilizer to one liter of water, I can calculate the expected ppms from: N, P, K;  Ca, Mg, S; (Fe, Zn, Cu... a few more listed on the box). That actually measures fairly accurately going through the circuitous conductivity->idealized-PPM of a reference solution

So, for awhile I've been thinking: why can't I calculate the idealized conductivity N, P, K (et. al.). I know the PPMs I'm adding of each.

I feel like I'm turning the concept 180 degrees. Instead of measuring conductivity (converted into idealized PPMs of a reference standard), the fertilizer product's label becomes the "ideal." I can calculate the resulting PPMs of a weight of fertilizer dissolved in 1L.). Why can't I know the expected microsiemens of that?

To me, it seems logical. I'm sorry if your explanation didn't sink in the way it should. I got caught on the opening "If..." In what way don't I know the breakdown of my $100?

I guess, coming at it 180 degrees: I've got a meter telling me my suitcase weighs $100 based upon an average weight of $1, $5, $10, $20. (Let's pretend each demonination has a unique weight). If I know I put 18 $1, and 5 $5, and 2 $20, and 1 $10. I could measure the weight more accurately and say that I actually have $98 in my suitcase?

I know the quantities of each denomination added to the water. I just don't know what they each weight (in terms of conductivity). I have to go through an abstracted conductivity->ppm (of a ideal) measurement.

[az2008]

Googling randomly led me to this page. About 2/3 of the way down that page, it reproduces a table from the CRC Handbook of Chemistry & Physics which shows common fertilizer compounds and the millisiemens they produce as percentages of a solution.

On the one hand: that suggests there is a way to calculate this. For example, it shows magnesium sulfate (MgSO4). I know from the last hurdle (5 years ago), that is 9.86% Mg, 13.01% S. (One gram/L produces 98.6ppm & 130.1ppm, respectively).

It seems like there would be some way to know how much the Mg contributed to the conductivity (and how much the S) did. It must get down to that level, I think(?).

On the other hand: It appears strength & conductivity are not linear. Using the epsom salt again: at 1% strength it has 7.6 millisiemens. You'd think a 10% strength would be 76. But, it's only 42.7.

So, apparently it's not as simple as I thought. If there were a way to calculate it (even if it were non-linear), I think I could do something with it. But, maybe it can't be calculated. (Maybe that book determined those values by making those dilutions, and measuring them.).

[Borek]

why can't I calculate the idealized conductivity N, P, K (et. al.). I know the PPMs I'm adding of each.

Conductivity depends on the ions present, not element. N can be present as NH₄⁺ and NO₃^-, each has a different conductivity. But it can be also introduced as an urea (CH₄N₂O) which is non ionic and doesn't add to the solution conductivity.

Using the epsom salt again: at 1% strength it has 7.6 millisiemens. You'd think a 10% strength would be 76. But, it's only 42.7.

This largely depends on the concentration. For diluted solutions conductivity is nicely additive and depends linearly on the concentrations, 10% is a quite concentrated solution and many additional effects start to play roles.

While the theory behind conductivity calculation is reasonably simple (at least for diluted solutions) it requires a rather solid understanding of some basic general chemistry concepts. I feel like if there is one solid advice that I can give you it is to start from the very beginning - GenChem101.

I guess, coming at it 180 degrees: I've got a meter telling me my suitcase weighs $100 based upon an average weight of $1, $5, $10, $20. (Let's pretend each demonination has a unique weight). If I know I put 18 $1, and 5 $5, and 2 $20, and 1 $10. I could measure the weight more accurately and say that I actually have $98 in my suitcase?

Unlikely. That is: in an idealized, theoretical case, knowing the ratio of of notes and coins - yes. In practice asking for 2% accuracy is way too much. You will never know the ratio of components (which doesn't actually mean NPK and so on, remember the N example above - by components I mean salts and non-ionic compounds present) precisely enough, plus all those effects making the dependence between conductivity and concentration non-linear will add another uncertainty.

[az2008]

While the theory behind conductivity calculation is reasonably simple (at least for diluted solutions) it requires a rather solid understanding of some basic general chemistry concepts.

Thank you for your time explaining it to me.

I think the roadblock for me will be that the fertilizer products list the atomic elements (N, P, K; then secondary Ca, Mg, S; then the trace elements like Fe, Zn). They don't usually (or completely) explain that Ca is provided as Calcium Nitrate (which supplies some of the N too). If I can't get to ions (or whatever's needed to calculate conductivity) from the amount of individual atomic elements (whose quantities are specified as "percent of weight" on the label), then it's probably not do-able for me.

It's possible fertilizers use standardized formulations (compounds) to supply those labeled "% of weight"  atomic elements. Maybe I'm just ignorant of that. Maybe that's what I need to investigate. If that were relatively consistent, maybe I could begin to unwrap the ions and calculation of conductivity in water.

This discussion has caused me to think what I should do is: dissolve 1 gram of a fertilizer product (in 1 liter water), measure the conductivity, and then have a "factor" to go between my calculated PPMs and the expected (observed) conductivity. Instead of one factor for all fertilizer products (like PPM pens have). I could have factors per product. Maybe that would be a starting point to relate my calculated PPMs to a measurable (sanity-check) result.

What this really boils down to (or, what I'm trying to improve upon) is how such info is communicated to someone else. My MgSO4 might weigh 0.7g/Tsp. I could say I use 1/2tsp (which should be 0.35g). But, someone else's might weight differently. Not everyone has a subgram scale. A PPM meter is more ubiquitous. It's easier to say "add enough to get 85ppm." But then, these meters are an abstraction. My meter may show 85ppm, but someone else's won't. That leads to microsiemens being a more direct and reproducible value. "This weight of epsom salt should produce 85ppm -- which should increase conductivity by 113uS (which is more accurately measured than PPM)"

If it's just me, my ppm pen, my fertilizer products, my subgram scale. The calculated and observed PPMs per gram would make sense. I would know what works for me, and how a product creates 10% more PPMs than calculated. The problem is communicating that to someone else. I was hoping to calculate expected conductivity ("add this much microsiemens of epsom salt."). But, if I observe the conductivity of the amount I know works well, that's about the same. That should be similarly reproducible.

Probably the best thing to do would be to keep a log (in a spreadsheet) of how much I add, and the resulting microsiemens. Then I could see how it differs when added to 200ppm water versus 500ppm water. (I could see how that non-linear growth of conductivity occurs.). I could develop my own predictive result from that history of observation.

[Borek]

"This weight of epsom salt should produce 85ppm -- which should increase conductivity by 113uS (which is more accurately measured than PPM)"

I don't think that's the case.

I am only guessing, but as I wrote earlier, my bet is that the ppm meter is just a conductivity meter, with the display scaled not in μS, but in whatevertheyareintendedtomean ppm units. So you won't get any better accuracy, just a better defined property of the solution.

[az2008]

"This weight of epsom salt should produce 85ppm -- which should increase conductivity by 113uS (which is more accurately measured than PPM)"

I don't think that's the case.

I am only guessing, but as I wrote earlier, my bet is that the ppm meter is just a conductivity meter, with the display scaled not in μS, but in whatevertheyareintendedtomean ppm units. So you won't get any better accuracy, just a better defined property of the solution.

You're absolutely right. This is a confusing topic because when I talk about PPMs, there are

1. "PPMs I calculate from the fertilizer label (% of weight)".

That's accurate within the limitation of the label being merely a "minimum guarantee." (This topic could also be limited to the US. I've read that other countries may list fertilizer content as % of volume. I've never looked into that topic.).

2. "PPMs as reported by a meter."

I agree with you. They're all different. Even if they have the same scale they can be different. I have 4-5 pens and often measure a nutrient solution with all of them (and compare how they measure each fertilizer product added).

There are some fairly inexpensive PPM pens that display microsiemens. One is "Membrane Solutions" (ASIN: B07R58VQWJ on Amazon for $10 USD). I've been using that for over a year, and it's a real bargain. It displays the usual confabulated "PPMs." But, press the "shift" button, and it displays microsiemens. So, I can see how it's scaling conductivity into a "PPM" representation. It can be calibrated. I use 342ppm NaCl. Really awesome for $10. It seems to be stable, doesn't require recalibration too often.

I may be able to calibrate it to 1000ppm NaCl. I haven't tried yet. I have a 2nd pen on the way. I'm going to calibrate it to 1000ppm (if possible) and then see how the two pens compare from weaker to stronger solution. (I.e., I start with 150ppm water. I add potassium sulfate to get 60ppm (calculated).). After adding some various fertilizer, the solution might be 650-750. I'd like to see how the same meter calibrated to the low & high range reads low and high solutions. I'd like to graph that, and see how the two together might provide more accurate readings at the mid point.

I also have a HM Com-100 ($55 USD). It displays microsiemens too (as well as both .5 and .7 scales). It is calibrated to the same 342 reference solution. The $10 "Membrane Solutions" tracks that pretty well until the 550-650 range, then either the inexpensive meter under-reports. Or, the expensive one is over reporting. That's why I want to compare two of the inexpensive meters to see if one calibrated for 1000ppm (assuming it can be), becomes more accurate at higher levels. How the two together compare to the expensive meter across that 150 to 900 range I tend to measure within.

So, I'm thinking I should be able to mix 1g/L potassium sulfate, measure the conductivity, and create my own "scale" to match the calculated PPMs (415ppm K; 170ppm S).

I calculate those now, and then expect to see those PPMs using the meter. But, as you said, that's hocus-pocus. I think it would be more accurate to measure conductivity, and maintain my own "calculated-ppm to conductivity" scale. My goal is to stop reading PPMs with a meter (since that's nonsense). I think the best I can do is say "I know this product creates x ppms when mixed 1g/L -- based upon it's 'guaranteed analysis.' 1g/L produces x microsiemens." From that, I could measure the microsiemens of adding an amount of that product, and convert that to ppms (myself, not having a meter display ppms to me).