Farm Operations Management
A straight fertilizer nutrient solution recipe is not a fixed answer, it is a blueprint you rebuild
Articles for Farm Operations Managers
You weigh out the amounts in the recipe, dissolve the straight fertilizers in order, measure the EC, and it lands right on your set point. No step was skipped, and the numbers check out. If that is the case, the contents must be right too — it is natural to think so.
But all EC guarantees you is the total amount of dissolved ions; it says nothing about the breakdown. Even when the recipe is correct and the EC checks out, the contents of the tank still drift, little by little, away from the design. This article is about nutrient solution mixing on the premise that it does drift.
Do not treat the recipe as a fixed answer
In summer, white deposits start to collect at the bottom of the tank. Or the leaves look a little different before harvest than they did last year. Small signs like these can stop your hand for a moment.
In the leafy-greens operations I have seen, it is not rare that the people who mix their nutrient solution from straight fertilizers keep using, indefinitely, the very first recipe someone handed them. They mix potassium nitrate, calcium nitrate, and monopotassium phosphate themselves. That is fine while it keeps running, but one day it nags at you: “is this really still good as it is?” The trigger is usually some trivial change like that.
You are careful to separate the calcium source from the phosphate and sulfate sources into different concentrated tanks. Even so, you cannot quite tell whether that white precipitate at the bottom is the same story. You have a sense that it precipitates more easily when the water temperature rises. But you use the same recipe all year round, and it feels like those two things do not line up. Have you had that experience?
At the root of this “the seasons move but the recipe does not” gap is the habit of treating the recipe as a fixed answer you set once and never touch again. In reality, how easily things precipitate and how much the crop needs both shift with the water temperature and the conditions of the moment. The recipe is not a fixed value to memorize; it is a blueprint you review at the turning points and rebuild when the premises break down. That will not send your yield through the roof. But it does help you cut wasted resources and head off deficiencies and precipitation before they happen.
Where this matters, and where it does not
Let me be honest up front about where this “rebuild it” story matters. The effort of rebuilding is worth it where these overlap: closed recirculation that reuses the nutrient solution, leafy greens, and a summer when the water temperature climbs. Here the EC checks out yet chloride and micronutrients quietly fall to zero, and calcium phosphate settles out — real harm that shows up clearly in the actual measurements of the research we will look at later. In this territory I can say flatly that it is worth rebuilding a fixed recipe.
Conversely, in run-to-waste setups, in non-recirculating ones where you make a fresh batch every time, in small operations that only run occasionally, and in winter when the water temperature does not rise, a fixed recipe runs just fine. If you do not reuse it, the ions do not accumulate and skew; and if the water is cold, you are far from the precipitation zone. Most of the operations that have “run on a fixed recipe for ten years” are on this side. So start this article by figuring out which side your own operation is on. What follows assumes recirculation, leafy greens, and summer.
What the white summer precipitate really is, and how to tell them apart
Something white shows up at the bottom. The first thing is not to misidentify it.
Separating the calcium source from the rest is the right move. But what you should separate here is the calcium source from the “phosphate and sulfate sources,” not the often-repeated “separate calcium from magnesium.” What pairs with calcium and settles out is the phosphate (calcium phosphate) and the sulfate (calcium sulfate). In practice, the basic approach is to put calcium nitrate in Part B and gather the phosphate, sulfate, potassium, magnesium, and micronutrients into Part A, and never let the two meet while concentrated. And the white deposit that shows up at the bottom in summer is, among these, the combination with phosphate — calcium phosphate is what to suspect.
That said, the precipitate is not always calcium phosphate. You can get a read on what settles at the bottom or in the solution from how it appears. Even within what I have seen on site, the following breakdown was the most practical.
- Calcium phosphate: a fine, white-to-grayish-white precipitate. Tends to appear above pH 6.0, and is common in summer.
- Calcium sulfate: a white, crystalline precipitate. Tends to appear at low temperature and high concentration. The opposite direction from calcium phosphate, which appears on the high-temperature side.
- Iron precipitate: brown to reddish-brown. Tends to appear above pH 6.5, and when exposed to sunlight.
- Calcium carbonate: a white powder. Tends to appear when you use hard water or above pH 7.0.
In short, you can make an educated guess: white and fine means calcium phosphate; crystalline in a cold season means calcium sulfate; hard water with a higher pH means calcium carbonate; brown means iron. Since the temperature direction is reversed, do not confuse what appears in the cold with what appears in the heat.
The sense that precipitation gets easier as the water temperature rises is, for calcium phosphate, correct. Calcium phosphate is known to have a reverse-solubility property where it actually becomes harder to dissolve the higher the temperature, and its tendency to appear in summer fits with this. That is exactly why running the same recipe all year round is itself slightly out of step.
This is not armchair theory. In research that tracked the nutrient solution in closed-recirculation hydroponics, the precipitate that built up in the system was reported to be amorphous calcium phosphate with a calcium-to-phosphorus ratio of about 1.2 (see 1). That precipitate dragged down iron and manganese with it as it settled. So “phosphate and calcium that stay together while concentrated will precipitate out” is recorded as something that actually happens in recirculating solution. Note, though, that this was observed under a high-pH condition of 8.0 to 8.5, and it does not give a quantitative threshold for how concentrated it has to be before how many grams settle out.
Let me splice this together with the numbers from the field. Most operations keep the nutrient solution pH around 5.5 to 6.2 as they run. The research’s 8.0 to 8.5 is, in that sense, an extreme condition. But it is too soon to relax with “we run around 6, so we will not precipitate.” Calcium phosphate gets easier to form above pH 6.0, and iron precipitation also begins above pH 6.5. In other words, just above the field’s operating band, the moment you enter the 6.0s, you are standing at the entrance to precipitation. As a tendency, reading it as “the more concentration, water temperature, and residence time line up, and the higher the source water’s alkalinity drives the pH up, the more it appears” is also sound as a matter of ordinary chemistry.
The first thing to sort out is whether the precipitation is happening in the stock-solution tank or in the diluted solution after mixing. Stock solution, being concentrated, is more prone to precipitate. What is easy to miss here is that even if you separate the phosphate source and the calcium source into different tanks, once dilution brings them together they meet after all. So use the stock solution diluted by roughly 100x, and always dilute Part A and Part B separately before combining them — the cardinal rule is never to mix them directly.
There is some relief, though. After dilution the concentration drops sharply, so even at the same temperature you move away from the precipitation zone. It is enough to hold the rough guideline that the tendency to precipitate rises the more those three — concentration, water temperature, and residence time — line up. If it is dilute, even if they do meet a little, it will not finish solidifying while it flows past. So what is truly scary is somewhere that should be dilute but stays concentrated and lingers for a long time. You premix the diluted solution and let it sit overnight. The liquid stagnates at a fitting or a dead end in the delivery piping. Then summer’s water temperature rides on top. When those three overlap, precipitation forms locally even where you thought you had diluted it. What to look at is whether it is the stock-solution tank, or, even after dilution, a “place where it collects and does not move.” If it is the latter, the move is to suspect the premix time and the stagnation in the piping before the recipe. Put another way, precipitation is a problem you close out on the operations side — temperature, residence, dilution — rather than by rebuilding the recipe itself.
Even when EC checks out, the contents run dry
Now that we have a handle on precipitation, the next question is how to set the recipe’s numbers themselves.
In most operations, people build the recipe off a target EC (a rough measure of the overall concentration of fertilizer dissolved in the nutrient solution). You add the amounts the recipe specifies so the EC comes out about the same as last year, match it with an EC meter, and you are done. You do not calculate each ion individually or break out the nitrogen and potassium. You stand on the premise that if the EC checks out, the contents are roughly right too. I ran it that way myself at first.
But EC is nothing more than the sum total of all the dissolved ions. Even if the total is the same, the ratio of the breakdown is a separate matter. Whether nitrogen is high and potassium is low, or the other way around, the EC meter shows the same number. The skew in the contents does not register on an EC meter.
Furthermore, when you reuse the nutrient solution in a recirculating system, the crop does not absorb the ions evenly. Some ions quietly decline, alone, without riding the overall movement of the EC. Recirculate while keeping a fixed recipe, and you get a situation where the EC checks out yet only some of the ions run dry.
This shows up clearly in the actual measurements too. In research that tracked each ion in closed hydroponics, the concentrations of nitrate, sulfate, magnesium, calcium, and potassium matched the overall movement of the EC well. They are on the side you can follow with an EC meter. But phosphate, sodium, chloride, and micronutrients such as iron and manganese strayed from the EC’s movement. Chloride in particular declined continuously from day 12 onward, throughout the cultivation period, and was finally drawn down to nearly zero. Iron and manganese also fell to nearly zero in the latter half (see 1). The phenomenon that “the EC holds, yet the ions EC cannot see drain first” is not a matter of intuition; it is a movement that becomes visible when you measure the ions individually.
What I want to flag here is which one drains first. In the same research, the uptake rate also differed considerably from ion to ion, and potassium itself was in the fastest-absorbed group. But being absorbed fast and draining from the solution first are two different things. Because potassium is taken up in large amounts, replenishment kept pace, and as a solution concentration it correlated well with EC. What actually fell to zero was chloride, sodium, phosphate, iron, and manganese. So rather than betting on a single point — “potassium drains first” — it is more faithful to the source’s findings to frame it as “the group of ions EC cannot see quietly drain dry; potassium is one example to watch because it is absorbed fast.”
There are cases where that skew reached all the way to yield. When hydroponic lettuce was recirculated while the nutrient solution was held at a target EC, the above-ground weight dropped by roughly 20 to 40 percent (up to 36 percent in the experiment) compared with a control grown on freshly made nutrient solution. At that point the plant’s nitrogen, phosphorus, potassium, and iron concentrations were all low together: the EC-meter number held, yet the contents were deficient (see 2). Note that the main mechanism by which nutrients skewed in this study was that source-water calcium, magnesium, and bicarbonate accumulated and propped up the EC, and in their shadow the needed ions thinned out. It is not the simple story of potassium alone being absorbed first and vanishing. Either way, the situation where the premise “as long as the EC checks out, the nutrients are sufficient” breaks down is left behind as numbers.
Back-calculate from the ions, not from a target EC
So how should it really be set? The line of reasoning is: rather than deriving the recipe amounts from a target EC, decide each ion’s target concentration first, and back-calculate the straight fertilizer amounts from there. The unit you use here is mEq/L (milliequivalents per liter). Plants take up nutrients in ionic form, so aligning by an ion’s electrical reactivity in equivalents rather than by weight makes the design fit. Potassium (monovalent) and calcium (divalent) differ in reactivity by about a factor of two for the same weight, and that is where mEq/L pays off.
The sequence goes like this. First, from the crop and the stage, set the target concentrations of the major ions — nitrogen, potassium, calcium, magnesium — in mEq/L. Next, fill them in starting from the straight fertilizers that leave the least ambiguity.
Concretely, let me work out the calcium nitrate amount using the Yamazaki formulation, widely used for lettuce, as an example. The Yamazaki formulation sets calcium at 2 mEq/L. Calcium nitrate (Ca(NO3)2.4H2O) has a molecular weight of 236.1, and since calcium is divalent its gram equivalent is 40.1 / 2 = 20.05. The amount needed per 1,000 L is target concentration x gram equivalent x molecular weight / atomic weight, so 2 x 20.05 x 236.1 / 40.1 = 236.1 g. In other words, for 1,000 L of nutrient solution, calcium nitrate is fixed at 236.1 g.
There is one crux to the back-calculation here. The moment you add calcium nitrate, nitrate-form nitrogen comes in along with the calcium. If, say, 4 mEq/L of nitrate comes in here and your target nitrate is 10 mEq/L, you fill the remaining 6 mEq/L with potassium nitrate — and so on: each time you fill one ion, you subtract the other ion that came in with it from the next calculation. Calcium is essentially fixed by calcium nitrate and magnesium by magnesium sulfate, so place those first; match the leftover nitrogen and potassium with potassium nitrate; and add phosphate and micronutrients last.
Then the EC follows, as a result, after the fact. EC is not the target you go set first; it is the tool you use last to check “has it drifted.” That said, hand-calculating every ion every time is not realistic in the field. In practice it is common to calculate with a nutrient program tool or a spreadsheet, and this site distributes a free tool that can calculate straight fertilizers and pre-mixed fertilizers together.
[hydroponics] A dead-simple, easy-to-use nutrient program tool: SimpleFert
Measurement, too, is hard to do for every item every time. Start by getting a read on what EC cannot see and what drains fast — chloride, micronutrients, and fast-absorbed potassium — and look at just those. Starting there is the realistic move.
Move the recipe with growth stage and water temperature
If uptake skews, is it fine to keep running the same recipe right after final planting and before harvest? That question naturally comes up. Because what the crop absorbs changes by stage, yet the recipe alone can stay frozen.
As the crop moves from vegetative growth toward fruit set and harvest, the direction of its demand does shift. For fruiting vegetables, it is said that during the leaf-building phase it leans toward nitrogen, and as the plant starts setting fruit the relative weight of potassium rises (since PFAL leafy greens are my main field and I have not watched fruiting vegetables firsthand, I leave this as something within what I have heard). Even so, this is not something you tinker with daily. Reviewing it only at the turning points is enough: right after final planting, at peak growth, and from fruit set to before harvest.
For the cue to switch, rather than deciding it mechanically by calendar, the realistic approach is to read the shrinking of the drain-prone ions together with how the crop looks. A drain-prone ion has fallen, and on top of that the crop’s appearance has started to change. You move when both of those line up. In practical terms, an easy-to-handle way is to take nutrient solution analysis, compare with last time, and accumulate the adjustment “nudge up next time any nutrient that dropped a lot, and lower any nutrient that barely changed” in steps of around 10 percent. The knack is not to swing big all at once; rather than aiming for the perfect formulation, a posture of converging while watching the response is what lasts in the field.
There is one thing I want to put down honestly. The effect of moving the recipe only kicks in once the environment — water temperature, variety, light — is in place. If the environment is poor, no matter how finely you move the recipe, the effect is limited. So treating dynamic recipe design as a defensive adjustment that heads off deficiencies and precipitation, rather than a move that sends yield soaring, is closest to my felt sense within what I have seen in leafy greens.
This “only kicks in once the environment is in place” property has been confirmed repeatedly in research too. In experiments looking at the relationship between EC and yield in hydroponic lettuce and peppers, the optimal EC value itself moved with the combination of season, variety, and water temperature, and could not be pinned to a single value by EC alone. There is even a report that what most governed yield was the combination of cultivation season and variety (see 3, 4, 5). Furthermore, there is an experimental result that holding the nutrient solution’s maximum temperature to around 26 degrees C made the growth drop that should have appeared at higher EC nearly vanish (see 5). Conversely, this means that while the temperature is under control, a bit of recipe drift is unlikely to surface, and it is precisely in summer, when that temperature control slips, that the recipe side starts to matter. But this is about the EC level — that is, the total salt load. Temperature can only paper over so much; it cannot hide the skew in the ion ratio itself. This overlaps with the framing of the trigger condition: what you should move is not constantly, but when the premises break down.
Signs that catch a drifting recipe early
The recipe has drifted. Or a particular ion has shrunk. To catch that early, what should you look at? Granting that you have an EC meter, are there other signs you can read as “if this value moves this way, it is a yellow light”? That is the question here. In the leafy-greens operations I have seen too, even when people watched EC and pH, they did not know what to watch beyond that, and ended up noticing the change in leaf color after the fact — that seemed common.
The earliest is the habit of narrowing to one or two drain-prone ions and watching them at the same point. Inspect the items you have a read on — chloride, micronutrients, fast-absorbed potassium — at a fixed point, by simple means if need be. Even if you cannot measure every item every time, just this surfaces the EC-invisible depletion first.
pH has its uses too. Whether the nutrient solution’s pH creeps up or down over time — that direction is explained as an indirect reflection of whether the crop absorbs more nitrate or more ammonium, that skew. But I do not recommend casting this as the lead actor in “an early sign of the contents skewing.” In operations with automatic pH control, the equipment cancels out the direction, so the “direction it moves” is itself invisible. It also moves with the source water’s alkalinity, dissolved carbon dioxide, and microbial activity, so multi-factor noise rides on it. What reference [6] shows is, if anything, pH’s limitation: “EC and pH cannot tell the individual ion species apart.” So pH fits best positioned as merely one supporting piece of evidence backing up the fixed-point measurement — one thread when you lay out and read the discrepancies among several values, such as “the EC holds yet the way the pH moves has changed from before” or “the EC is slow to recover relative to how much I am replenishing.”
This point that “EC and pH alone cannot capture the imbalance of individual ions” is taken up in research oriented toward building in ion-selective electrodes too, in the form that management which does not differentiate EC and pH cannot tell the individual ion species apart (see 6). But that simple individual measurement has its own quirks. An error is reported where the ion electrode reads potassium low, and the correction for it makes the actual mixed solution about 40 percent more concentrated (see 7). Calcium too reads low at the electrode when the background liquid differs between calibration and measurement, making the mixed solution about 30 percent more concentrated (see 8). So “measure individually because it does not register on EC” is the right direction, but do not take the simple-measurement value as correct on the first read either; the posture that fits is to read the direction it moved across several readings.
Let me add one more thing, a mechanism for catching mistakes. Weighing, once you have dissolved it, cannot be told by sight. Preparations before the work pay off: make a checklist of what has been weighed and tick each one off, color-code the Part A and Part B containers. When an error happens anyway, a safe line to draw is: within 15 percent of the target amount, add the shortfall or correct by dilution; beyond 15 percent, do not force a fix — discard and remake. When you add iron, always do it on the Part A side and stir immediately to prevent oxidation. Chelated iron differs in the acidic range it holds up in — Fe-EDTA at pH 4.0 to 6.5, Fe-DTPA at 4.0 to 7.5, Fe-EDDHA up to 4.0 to 9.0 — so if the source water’s pH runs high, the move is to choose a more stable chelate (the more stable ones are more expensive).
The value of straight fertilizer mixing is not cheapness
The motive for reaching into straight fertilizer mixing is usually “I want it cheaper than commercial pre-mixed fertilizer.” So, as intended, does switching to straight fertilizers really lower the cost? If you have read this far, you must be wondering whether, given the added effort of measuring and rebuilding, you come out ahead even if the fertilizer bill drops.
Honestly, I cannot say straight fertilizers are “always cheaper.” The unit price of the raw material itself does tend to drop. But in exchange, the labor of mixing, the cost of measurement, the inventory risk of holding many kinds of straight fertilizer, and the portion that gets discarded before you use it up all ride along against the fertilizer savings. So you cannot judge whether you come out ahead by looking at the fertilizer bill alone. On top of that, the profit and loss flips with scale and operation. If you make small amounts only occasionally, the convenience of a commercial mix can win; if you run large volumes and have a setup where you can move things yourself, straight fertilizers start to pay. Whether it is cheaper depends on the conditions, and you cannot settle it by looking at that alone.
The real value of straight fertilizer mixing lies elsewhere. The view is that it lies in the freedom to move the contents to fit the crop and the stage, more than in cheapness. Measure by cost alone and the advantages laid out here — being able to correct the skew, being able to avoid precipitation — drop out wholesale. In fact, there is research suggesting that quantitative nutrient solution management, which builds the application rate to fit each ion’s uptake, has an advantage in resource efficiency over managing by EC alone (see 9). But this stays at “suggesting”; it does not go as far as squaring off EC management against yield or cost and quantifying which wins. So I cannot say “straight fertilizer mixing always wins on both yield and cost,” but I can say “the freedom to converge the contents toward the crop remains as room that is hard to gain under management on EC alone.” A precise estimate that folds in electricity and labor costs can be carved out as a separate story, but here I will keep it within the scope of recipe design.
Finally, there is one point I want you to hold onto. A recipe you can guarantee will never precipitate — “follow this table and it will never settle out” — probably does not exist. Because once the source water’s hardness, the water temperature, and the tank’s residence time change, the conditions for precipitation move too. So this is not about memorizing specific numbers. First, check the current values at your own facility once, for yourself: the source water quality, the water temperature, and which ions drain easily. On that basis, decide one or two candidates to try in the next cycle. And when you do put out specific numbers, do not forget to look at target EC, ion ratio, water temperature, and target crop together as one set of premises.