Water, EC, and pH management in Bangkok hydroponics: what the logged data actually shows

TL;DR

Every number in a standard hydroponic guide -- the baseline EC of tap water, how fast pH drifts, how much water a reservoir loses per day, when a pump starts cavitating -- was measured somewhere that is not Bangkok. Across nine logged grow cycles in a Bangkok apartment, those numbers were wrong in the same directions, for the same reasons, consistently enough to document. This article maps what the sensor data actually shows across water quality, EC management, pH behaviour, and reservoir dynamics in a tropical grow environment, and links to the detailed logs behind each finding.

The problem with numbers that were not measured here

The first season of running a Bangkok hydroponic setup produces a specific kind of confusion. Everything is calibrated, the nutrient formula is correct, the EC target is within the recommended range -- and the crops are inconsistent in ways that no individual variable seems to explain. Tip burn appears in dry season without an obvious cause. pH corrections that held for two days in November need repeating every morning by March. The reservoir that lasted three days between top-ups in October is empty before 48 hours are up in April. The setup guide has not changed. The environment has.

The pattern only becomes legible once you start logging time-stamped data across multiple cycles. Not because the underlying mechanisms are complicated -- they are not -- but because the variables that drive them (Bangkok tap water chemistry, seasonal vapour pressure deficit, ambient reservoir temperature, roof tank mineral accumulation) are local. No guide written for a temperate climate accounts for them, because no guide written for a temperate climate was measured in Bangkok.

The input you treat as fixed is the one that moves the most

Bangkok tap water EC is not a fixed input. It moves by 0.3 to 0.4 mS/cm across the calendar year, driven by a mechanism that most growers have never been told about: the Bangkok Metropolitan Waterworks Authority blends output from two treatment plants drawing from the Mae Klong and Chao Phraya river systems, and the dissolved mineral load of those rivers shifts with rainfall and reservoir levels. In dry season (November to April), lower river flow concentrates dissolved minerals and allows salinity intrusion from the Gulf to push further upstream. In wet season (May to October), heavy rainfall dilutes the source water and pushes salinity back downstream. The tap at your sink reflects this seasonal chemistry -- not dramatically, but enough to matter. A grower who mixed to a target of 1.4 mS/cm in December and never remeasured the tap in June is running 0.3 to 0.4 mS/cm above that target at the root zone without knowing it. Tip burn attributed to calcium deficiency, pale wet-season growth attributed to light reduction, inconsistent results across the year that never resolve -- these are the crop-level symptoms of treating a seasonal variable as a fixed zero. There is also a secondary mechanism specific to Bangkok apartment growing that almost no guide mentions: the rooftop tank. A large fraction of Bangkok condos store mains water in a rooftop cistern before it reaches the tap. Those tanks accumulate mineral deposits over time, which means EC at your tap does not perfectly track treatment plant output. A recently cleaned tank reads lower than the mains average. A tank that has not been descaled in three years reads higher. The only number that matters is the one your pen reads right now, in your building, before each reservoir mix.

Across 24 months of pre-mix tap measurements at our On Nut grow location, dry season readings averaged 0.53 mS/cm. Wet season readings averaged 0.24 mS/cm. Within those seasons, individual readings varied by as much as 0.2 mS/cm depending on recent rainfall and suspected tank condition. The seasonal pattern is reliable. The daily number is not. Checking once per season is still under-accounting for the variation. The correct cadence is per batch, every time. (Plant Mood grow log data, 2023 to 2024.)

Tap water is a stable, known baseline. You measure it once, account for it in your nutrient calculation, and move on. The variable that changes is the nutrient you add, not the water underneath it.

What the logs showed across nine cycles

Dry season tap water at our On Nut location averaged 0.53 mS/cm (November to April). Wet season averaged 0.24 mS/cm (May to October). The lowest single reading was 0.15 mS/cm after an extended heavy rain event in August 2023. The highest was 0.67 mS/cm in March 2024. A grower targeting 1.4 mS/cm who does not adjust for the dry-season tap baseline can overshoot by 25 to 33 percent before adding a single drop of nutrient. In four logged lettuce cycles, tip burn between day 16 and day 20 correlated with pre-mix tap readings above 0.55 mS/cm -- the reservoir was mixed to target, but the starting point had shifted. (Plant Mood grow log data, 2023 to 2024.) Full data: Bangkok tap water baseline EC: what changes between dry season and wet season.

Bangkok tap water EC swings 0.3 to 0.4 mS/cm between dry and wet season -- and within-season variance is large enough that monthly checks are insufficient

Measure tap EC before every reservoir mix, not once per season. Calculate available headroom (target EC minus tap EC) before adding nutrients. The label dose assumes a starting point that changes with every month of the Bangkok calendar.

A 14-day paired logging trial ran two identical 60L DWC reservoirs, one held at 31 to 33 degrees Celsius and one at 24 to 26 degrees Celsius using a chiller. The hot reservoir drifted pH upward at an average of 0.31 units per day. The cool reservoir drifted at 0.09 units per day. The hot reservoir required pH correction on 11 of 14 days. The cool reservoir required correction on 4 of 14 days. The driver is root respiration and microbial metabolic activity, not the direct chemical effect of temperature on water equilibrium -- which matters because the drift rate changes with crop stage and root mass, not just the thermometer. Day 5 to day 10 produced the highest single-day drift at 0.48 units on day 8, corresponding to peak root mass expansion. (Plant Mood grow log data, June 2024.) Full data: pH creep rate at different water temperatures: 14 days of logged data.

A hotter reservoir drifts pH upward at more than three times the rate of a cool one -- and the driver is biological, not chemical

A Bangkok DWC reservoir above 30 degrees Celsius during active crop growth cannot be managed with a once-daily pH check. The morning reading is the overnight low -- the least stressed moment of the day. At 0.31 units per day average drift, the reservoir peaks an average of 0.43 units above the morning reading by 3pm. That afternoon excursion falls in the pH range where iron and manganese availability declines.

Across multiple Bangkok tower cycles, basil depleted EC at an average peak rate of 0.18 mS/cm per day (days 10 to 18), lettuce at 0.11 mS/cm per day (days 6 to 12), and kale at 0.07 mS/cm per day across its full cycle -- a ratio of roughly 2.6 to 1.6 to 1.0 at respective peaks. On ambient days above 33 degrees Celsius, the basil-to-kale gap widened from 0.11 to 0.15 mS/cm per day as heat amplified the depletion difference. A reservoir corrected to serve basil peak demand over-concentrates kale. A schedule optimised for kale starves basil during its mid-cycle spike. Using individual ion test strips alongside EC readings showed the calcium-to-potassium ratio drifting measurably toward elevated calcium by day 10 in mixed runs, consistent with crops selectively removing different ion ratios. By day 14, the solution no longer matched the intended formula for any crop. (Plant Mood grow log data, 2023 to 2024.) Full data: EC depletion rate by crop type on the same tower: lettuce vs basil vs kale.

Lettuce, basil, and kale deplete EC at different rates and different crop phases -- a shared reservoir cannot serve all three simultaneously

Total reservoir solution replacement at day 10 to 12 in a mixed tower -- not top-ups alone -- partially resets the ionic composition drift. A single EC correction schedule cannot be correct for all crops at all cycle phases.

Across 47 logged top-up events in DWC and NFT systems, equilibration time (defined as less than 0.02 mS/cm drift across three consecutive 60-second readings) ranged from 8 to 22 minutes. The average divergence between the immediate post top-up reading and the stabilised reading was 0.19 mS/cm. In 33 of 47 events the immediate reading was lower than stabilised; in 14 it was higher -- direction depends on sensor placement relative to the top-up inlet, making the error unpredictable without knowing system geometry. Events above 30 degrees Celsius solution temperature took an average of 17.4 minutes to stabilise. Events below 28 degrees Celsius took 10.2 minutes. Temperature explained more variance in equilibration time than top-up volume. Growers who made a second addition within 5 minutes of the first produced overshoot events in roughly 4 of 5 attempts. (Plant Mood grow log data, 2023 to 2024.) Full data: EC recovery curve after topping up a depleted reservoir.

EC readings after a reservoir top-up continue moving for 8 to 22 minutes -- and solution temperature is the stronger predictor of how long

Temperate-climate guides typically suggest a 10-minute wait after top-up. In a Bangkok dry-season reservoir at 30 to 31 degrees Celsius, 10 minutes is often insufficient. The correct wait is 15 to 20 minutes for DWC and a minimum of 10 minutes for NFT. Dose after equilibration, not before.

Photosynthesis removes CO2 from reservoir water during daylight hours, which reduces carbonic acid concentration and raises pH. In DWC lettuce reservoirs positioned near west-facing windows in Bangkok dry season, the morning-to-afternoon swing was 0.4 to 0.6 pH units on clear days. On heavy overcast days the same reservoirs showed 0.1 to 0.15 unit swings. Algae-contaminated reservoirs amplified the swing to 0.9 units. The morning reading is the overnight low after 8 to 10 hours of reduced biological activity -- the most reassuring number of the day and the least representative of where pH sits during peak plant growth hours. In three logged cycles where afternoon corrections were applied to photosynthesis-driven highs, the next morning reading fell below pH 5.9 every time. The pH-down collided with the overnight CO2 recovery the system was already going to make. (Plant Mood grow log data, 2023 to 2024.) Full data: pH vs time of day: does photosynthesis visibly shift reservoir pH on a sunny Bangkok afternoon.

pH rises 0.4 to 0.6 units between 7am and 3pm on clear Bangkok days -- and correcting the afternoon reading creates the crash you see the next morning

Test pH between 6am and 7:30am for any reading that will drive a management decision. Accept that afternoon readings will be elevated on clear days. Correct only when the morning baseline itself has shifted outside the target range.

Ultrasonic level sensor logs across multiple Bangkok tower cycles showed a full 28-head lettuce tower consuming 3.9 to 4.8 litres per day at canopy maturity in dry season -- against the 2 to 3 litres per day cited in most tower setup guides. Basil on the same hardware reached 5.6 to 6.1 litres per day at peak. Wet season reduced consumption by 25 to 40 percent across all crops, driven primarily by lower vapour pressure deficit rather than temperature. The acceleration window between day 14 and day 22 in a dry-season lettuce cycle saw daily draw increase at an average of 0.35 litres per day. An overnight consumption drop of more than 30 percent below the three-night rolling average appeared 36 to 48 hours before visible root zone stress symptoms in two separate cycles -- the sensor flagged the problem before the canopy showed anything. (Plant Mood grow log data, 2023 to 2024.) Full data: Daily water consumption of a full hydroponic tower by crop and season: ultrasonic level sensor data.

A hydroponic tower uses 60 to 90 percent more water per day at canopy maturity than the setup guide estimates -- and the draw accelerates precisely in the week before harvest

Size reservoirs to peak daily consumption (5 litres per day for dry-season lettuce, 6.5 litres per day for basil) not average consumption. Operate two top-up intervals: a relaxed one for days 1 to 13 and a compressed one from day 14 to harvest. The schedule that works in week one will fail in week three.

A 30-day level sensor log across Bangkok April 2024 in a non-climate-controlled apartment showed a 50-litre reservoir running mature lettuce needing top-up every 1.8 days on average. On four days where outdoor temperature exceeded 38 degrees Celsius and the grow room climbed above 34 degrees Celsius, daily draw exceeded 9.3 litres. A fixed two-day top-up interval -- which held through February and March without failure -- failed 6 times in April. Every failure correlated with outdoor temperatures above 36 degrees Celsius. Sealing all lid gaps with reflective foam tape reduced daily draw by approximately 1.4 litres on comparable temperature days -- a 20 percent reduction from a 20-minute intervention. The 5pm to 7pm window produced the highest hourly draw of the day due to thermal lag between outdoor peak temperature and grow room peak temperature. (Plant Mood grow log data, April 2024.) Full data: How often does a 50-litre reservoir need topping up in Bangkok April heat? 30-day level sensor log.

In Bangkok April heat, a 50-litre reservoir running mature lettuce required topping up every 1.8 days on average -- and on the hottest days drew over 9 litres

From April 1, switch to daily morning reservoir checks regardless of what the previous day showed. A two-day interval calibrated to March will fail repeatedly in April. Before April arrives, seal the lid. It is the highest return-on-effort intervention available before automation.

Across 34 logged low-level events with JSN-SR04T sensor data and simultaneous pump audio observation, audible cavitation onset occurred consistently when water depth above the pump inlet fell below 6 to 8 centimetres -- in 29 of 34 events. In 5 events driven by air stone turbulence near the inlet, the threshold reached 9 to 10 centimetres. The 5-centimetre safety margin cited in several guides describes where cavitation becomes severe, not where it begins. In 11 of 34 events, flow rate declined measurably before any audible change -- silent efficiency reduction that no grower without a flow rate sensor would detect. Across two pump lifecycles, cumulative cavitation exposure correlated with flow rate degradation: 7 logged events produced a pump ending life at 88 percent of rated flow; 19 events produced one ending at 71 percent. At April peak consumption above 9 litres per day, the buffer between an 8-centimetre alert threshold and actual pump air ingestion compressed to approximately 23 minutes. (Plant Mood grow log data, 2023 to 2024.) Full data: The relationship between reservoir water level and pump cavitation risk: what the JSN-SR04T data shows.

Pump cavitation begins when usable water depth falls below 6 to 8 centimetres above the pump inlet -- not at empty -- and the alert-to-action window compresses to 23 minutes in Bangkok April heat

Set low-level alerts at 8 centimetres above the pump inlet, not at the inlet level itself. In April, raise that to 10 to 12 centimetres. Mount pumps on a 3-centimetre riser -- counterintuitively, this reduces the cavitation threshold by improving lateral inlet flow geometry and eliminating the floor-levelness variable.

Across 14 logged H2O2 dosing events at 3 ml/L of 3%, TDS increased by 40 to 70 ppm within 5 to 10 minutes. The spike is caused by ionic byproducts from H2O2 decomposition interacting with chelated iron and organic matter in the reservoir -- not by any nutrient addition. In Bangkok reservoirs above 29 degrees Celsius, the spike resolved in under 35 minutes in 6 of 8 events, due to accelerated decomposition in warm water. Day 1 reservoirs (low organic load) showed median TDS increases of 18 ppm. Day 7 reservoirs showed 63 ppm from the same dose. Growers who check EC within 15 to 20 minutes of dosing and act on the reading risk reducing nutrients that were not too high -- and triggering lower-leaf deficiency symptoms 3 to 4 days later that get attributed to the root problem rather than the accidental underfeed. (Plant Mood grow log data, 2023 to 2024.) Full data: Does hydrogen peroxide affect EC readings? What the TDS sensor shows before and after.

Hydrogen peroxide produces a TDS spike of 40 to 70 ppm within 10 minutes of dosing -- and the spike is larger in older, organically loaded reservoirs than in fresh ones

Record pre-dose EC baseline before every H2O2 treatment. Wait a minimum of 90 minutes in cool reservoirs or 60 minutes in Bangkok reservoirs above 28 degrees Celsius before taking any EC reading that will inform a nutrient decision.

Plant Mood grow log data, 2023 to 2024

Why Bangkok specifically produces all of these deviations

The deviations documented across these nine cycles share a common root: tropical growing conditions change the rate, magnitude, and direction of chemical and biological processes in ways that temperate-climate guides do not account for -- because they were not measured there. Three mechanisms run through nearly every finding in this cluster.

The first is temperature. Biological reaction rates roughly double for every 10 degrees Celsius increase in temperature (the Q10 relationship, standard plant physiology). A Bangkok DWC reservoir at 30 to 32 degrees Celsius is running root respiration, microbial metabolism, and H2O2 decomposition at approximately twice the rate of a reservoir at 20 to 22 degrees Celsius. That acceleration is not an edge case for a Bangkok summer heatwave -- it is the default operating condition for most of the year in an unconditioned grow space. Every guide value that depends on a reaction rate (pH drift per day, H2O2 spike resolution time, EC equilibration after top-up) was calibrated in conditions that do not apply here.

The second is vapour pressure deficit. Bangkok dry season at 33 degrees Celsius and 55 percent relative humidity produces a VPD of approximately 2.0 kPa. A 31-degree Celsius wet-season day at 85 percent humidity produces 0.85 kPa. Plants transpire to cool leaf surfaces, and the rate is driven by how much the surrounding air can absorb -- which is what VPD measures. Tower water consumption differs by 25 to 40 percent between Bangkok dry and wet season not primarily because of temperature but because of the difference in atmospheric water demand. No setup guide that gives a single daily consumption estimate accounts for this, because the estimate was not measured across a full tropical seasonal cycle.

The third is source water variability. Bangkok tap water carries a seasonal mineral load that most growers treat as fixed. The seasonal EC swing is 0.3 to 0.4 mS/cm between peak dry and peak wet season -- real enough to shift nutrient math, small enough to miss without a meter. The building tank layer adds within-season variance on top of that. A grower without a calibrated EC pen measuring tap water before every mix is running a variable they cannot see, in both directions, for the full year.

The most instructive single example is the tip burn pattern documented in the tap water EC article. In four dry-season lettuce cycles, tip burn appeared between day 16 and day 20. The standard diagnosis is calcium deficiency. The standard intervention is adding cal-mag. Both assumptions were wrong: the tip burn was EC overshoot caused by mixing to a nominal target without accounting for the elevated dry-season tap baseline. Adding cal-mag to a reservoir already running above its actual target EC made the problem worse. The data that resolved the confusion was a pre-mix tap reading -- 20 seconds of work that none of the preceding troubleshooting had included. The mechanism was not complicated. It was invisible without measurement.

The compounding effect deserves mention. In Bangkok dry season, three of the mechanisms documented here can operate simultaneously: elevated tap water EC starts the nutrient solution above baseline; higher reservoir temperatures accelerate evaporation and concentrate EC further between top-ups; and pH drifts faster than a once-daily correction schedule can track, causing nutrient lockout during the afternoon excursion. A grower experiencing a difficult dry-season crop is likely dealing with all three at once, which is why single-variable troubleshooting rarely resolves it. The data layer exists to see all three -- tap EC per batch, reservoir EC twice daily, pH at 7am and 3pm -- and the cost of adding those measurements is minutes per day.

The measurement stack that covers all nine findings

Measure tap water EC before every reservoir mix

A logged tap EC number for every batch. Calculate available headroom (target EC minus tap EC) before adding nutrient concentrate. Adjust dose accordingly -- not the label dose, the headroom dose.

Bangkok tap EC varies by up to 0.2 mS/cm within a season and 0.3 to 0.4 mS/cm across seasons. A reading from two weeks ago is not a reliable proxy for today.

Log reservoir EC and volume at a consistent time before every top-up

A timestamped consumption log with outdoor temperature noted. After 10 days, the acceleration curve becomes visible and predictable.

Daily consumption rate is the only number that reveals the crop growth stage transition from low-draw to high-draw -- and in April, the number that tells you whether the top-up schedule needs compressing before the reservoir empties.

After any top-up, wait 15 to 20 minutes before taking a DWC EC reading, 10 minutes for NFT

An EC reading that reflects bulk reservoir concentration, not local concentration at the sensor position.

EC readings continue drifting for 8 to 22 minutes post top-up depending on reservoir temperature and pump circulation rate. The immediate reading is wrong in a direction you cannot predict without knowing sensor placement.

Test pH between 6am and 7:30am for any reading that will drive a management decision

A comparable, consistent pH baseline across days and cycles. Reserve afternoon readings for in-the-moment diagnosis of a suspected problem, not routine management.

Morning readings have a standard deviation of 0.11 pH units across clean-reservoir cycles. Afternoon readings on clear Bangkok days have 0.29 units -- driven by photosynthetic CO2 drawdown that is self-correcting overnight. Correcting the afternoon reading creates the crash you see the next morning.

Set low-level reservoir alerts at 8 centimetres above the pump inlet, not at the inlet itself

A cavitation alert that fires before damage begins. Raise the threshold to 10 to 12 centimetres in April when the alert-to-action window compresses.

Cavitation onset in our logged events occurred consistently at 6 to 8 centimetres above the inlet -- partial air entrainment from surface turbulence, not inlet exposure. In Bangkok April at 9 litres per day draw, the buffer between the 8-centimetre threshold and pump air ingestion is approximately 23 minutes.

After any H2O2 dose, record the pre-dose EC, wait 60 to 90 minutes, then take the reading

A post-dose EC reading that compares accurately to the pre-dose baseline. A difference of less than 0.2 mS/cm means no nutrient adjustment is needed from the treatment alone.

H2O2 decomposition produces a TDS spike of 40 to 70 ppm within 10 minutes of dosing. In warm Bangkok reservoirs the spike resolves in 35 to 50 minutes. A reading taken during the spike window will show EC higher than actual nutrient concentration.

The assumptions that make all of this invisible

Standard hydroponic guides give specific numbers: pH target 5.8 to 6.2, check daily, EC target 1.2 to 1.6 mS/cm, size reservoir for 2 to 3 litres per day daily consumption, set pump alert at the inlet level. Every one of these numbers was generated in conditions that differ from Bangkok in measurable, documented ways. The pH range is correct but the check frequency is not adequate for a 30-degree Bangkok reservoir. The EC target is reasonable but the nutrient dose to reach it changes with the tap baseline, which changes every month. The reservoir consumption estimate is accurate for week one, not week three. The pump alert level is where cavitation becomes severe, not where it begins. The numbers are not wrong -- they are calibrated for the wrong environment.

Treating guide values as calibrated for Bangkok

Tip burn attributed to calcium deficiency. Pale wet-season growth attributed to light reduction from cloud cover. Slow basil on a mixed tower attributed to spacing or root competition. Intermittent iron deficiency symptoms in plants with a morning pH reading that looks fine. Each of these is a real diagnosis that appears in grow guides as a categorical answer. Each of them, in our logged cycles, was caused primarily or partly by a water-layer variable that was not being measured at the time: tap EC overshoot, wet season underfeeding, basil EC depletion during its mid-cycle spike, afternoon pH excursion above the iron availability threshold. The diagnostic sequence matters: measure the water before changing the nutrients, the light, or the seeds.

Diagnosing crop problems without measuring the water they grew in

Every published Bangkok water quality figure -- from the Metropolitan Waterworks Authority or otherwise -- describes the mains supply at the point of treatment, not at the point of use. In most Bangkok condos and shophouses, water passes through a rooftop cistern and sometimes a ground-level tank before it reaches the tap. Those vessels accumulate mineral deposits and sediment over time, which can add or reduce EC independently of the seasonal mains variation. A grower in a building with a recently cleaned tank and a grower two floors above with a tank that has not been descaled in three years are not working with the same tap water, even if they live in the same building on the same day. The waterworks supply data cannot account for this. The EC pen in your hand can.

The building tank as an invisible variable between the mains and the grow

A top-up schedule that held without failure through February will begin failing in March and fail consistently in April for any Bangkok grower running a reservoir without automation and without daily monitoring. The failure is not random. It is structurally predictable: VPD climbs with the dry season, transpiration rate follows, the crop is simultaneously approaching peak canopy consumption, and the reservoir that lasted three days in October is running dry in under two days. No single observation from a cooler month tells you this is coming. The sensor log, with its daily consumption curve and season-over-season comparison, does.

Assuming a fixed top-up schedule is a solved problem

What a Bangkok grower who has read this should change this week

Every Bangkok hydroponic grower, regardless of system size

Add a tap EC measurement to every reservoir mix routine. Keep a calibrated EC pen next to the water source. The measurement takes 20 seconds. Log the number alongside the date and batch. After five batches, you have data on how your building's tap EC is behaving. After ten, you have enough to see the seasonal pattern beginning.

Nutrient dose is calibrated to actual starting water EC, not an assumed fixed zero. Tip burn and pale growth caused by seasonal EC mismatch become diagnosable before they cost a crop.

Growers running DWC or tower systems in unconditioned Bangkok spaces

Shift pH testing to 6am to 7:30am as the management reading and add a second check at 3pm during the first week of any new crop cycle. The 3pm reading, compared to the morning baseline, tells you the amplitude of the diurnal swing in your specific system under current seasonal conditions. Once you know that amplitude for your setup, you can stop treating the afternoon elevation as a problem to correct.

Fewer unnecessary pH corrections. No more next-morning pH crashes caused by correcting a photosynthesis-driven afternoon high. A calibrated understanding of what the diurnal swing looks like in your system versus a genuine drift event.

Growers approaching April or currently running through it

Seal every gap in the reservoir lid this week with reflective foam tape -- cable entry points, net cup rims, probe holes, lid edge gaps. It takes 20 minutes. Based on our April log data, this single intervention reduces daily draw by approximately 1.4 litres on peak heat days. For a 50-litre reservoir drawing 8 litres per day in April heat, that is equivalent to gaining roughly four hours of additional buffer before the next top-up is needed.

Reduced daily water loss through direct evaporation. Slightly lower grow room temperature at peak afternoon. A longer buffer between top-ups in the month that most commonly produces reservoir dry-out events.

Growers who have experienced unexplained crop variation across seasons

Go back through any grow logs that recorded final reservoir EC but not tap EC. Using the seasonal averages from this pillar as proxies (0.53 mS/cm dry season, 0.24 mS/cm wet season), estimate the tap EC contribution for each past batch and recalculate the actual versus intended nutrient EC. Map the result against any yield or quality anomalies you noted. In most cases, the seasonal mismatch will be visible in the retrospective data.

Reclassification of past failures previously attributed to seeds, light, or nutrient formula. A clearer picture of which variables were actually driving variation -- and which ones are now measurable going forward.

Growers using hydrogen peroxide as a recurring reservoir treatment

Log pre-dose EC for every treatment event. Set a 90-minute timer after dosing before taking any EC reading that will inform a nutrient adjustment. If the post-dose EC reading is more than 0.2 mS/cm above the pre-dose baseline after 90 minutes, the cause is not the H2O2 -- check for probe fouling or a low-volume concentrated reservoir.

No accidental underfeeding events caused by responding to a TDS spike as if it were a genuine nutrient excess. A documented pre-dose and post-dose record that distinguishes H2O2 treatment effects from actual reservoir chemistry changes.

Logging the date alongside every measurement. An EC reading without a date is a data point. Ten EC readings with dates are a pattern. The seasonal tap EC swing, the consumption acceleration curve, the pH diurnal amplitude by sky condition -- none of these are visible from a single reading. They are visible from a sequence of readings taken at consistent times and written down. The growers who have this data do not need to guess why April is different from October or why the lettuce looked fine in November and stressed in March. The answer is in the log.

One thing to do before the next reservoir mix

Every finding in this article required the same first step: measuring something that was previously assumed to be known. The tap EC before a reservoir mix. The pH at 3pm as well as 7am. The reservoir level at a consistent time each day before topping up. None of these measurements require expensive equipment or a reconfigured grow space. They require the habit of writing a number down before acting on an assumption. Start with the tap. Before the next reservoir mix, dip the EC pen in the tap water, read the number, and write it next to the date. That number will tell you more about what your crop experienced this season than any troubleshooting article will.