Hydroponic water, EC, and pH management in Bangkok: what the sensor logs actually show

Key takeaways

In a passive DWC system in Bangkok conditions above 31 degrees Celsius, pH rose from 6.1 to 7.6 across 14 days at a mean rate of 0.11 units per day, accelerating to 0.16 units per day after day 8 as root mass increased. Plant Mood grow log data, 2024.

EC rose 64% across 7 days in a Bangkok hydroponic tower without any nutrient addition, driven by evaporation concentrating the solution. A subsequent plain-water top-up dropped EC from 2.3 to 0.9 mS/cm, below the minimum effective range for leafy greens. Plant Mood grow log data, 2024.

A reservoir sitting on a concrete floor in a Bangkok room at 30 degrees Celsius peaked at a mean daily high of 31.4 degrees Celsius, versus 28.1 degrees Celsius for the same reservoir on a 5cm timber riser in the same room. The 3.3-degree gap reduces dissolved oxygen saturation by enough to produce root browning from week two onward. Plant Mood grow log data, 2024.

Sunflower hydroponic germination held above 90% at EC 1.2 mS/cm and dropped to 51% at EC 2.0 mS/cm across tracked Bangkok cycles. The failure signal is hull adhesion visible on day 2, not the germination count at day 3. Plant Mood grow log data, 2024.

Root-driven pH drift in a hydroponic tower increased from 0.08 units per day in week one to 0.31 units per day in week four as plant load increased, with Bangkok dry season conditions producing drift rates approximately 35% faster than equivalent wet season cycles. Plant Mood grow log data, 2024.

Bangkok dry season daily reservoir volume loss averaged 4.1% of reservoir capacity per day across monitored NFT systems, roughly 1.8 times the wet season rate. A top-off volume calibrated in October will structurally under-compensate by March. Plant Mood grow log data, 2024.

The single cheapest intervention that changes the EC and pH management equation in a passive hydroponic system is an aquarium air stone at 1 to 2 litres per minute. In comparison cycles it reduced daily pH creep rate by 36% and extended post-correction pH stability from 36-48 hours to 54-60 hours. Cost in Bangkok: under 150 baht.

TL;DR

Every guide on hydroponic water management gives you numbers. Almost none of them were logged in Bangkok. The data from our grow cycles shows pH, EC, and reservoir temperature all behaving differently here than temperate-climate advice predicts, and the differences are not small. pH creeps faster, EC concentrates faster, floors heat reservoirs from below, and the correction timing that works in October fails in March. This article pulls the sensor data from six cluster articles into one place and explains what it means for how you actually run a system in this city.

The week three tray inspection that changed how we log everything

The lettuce looked fine from above. The reservoir was topped up on schedule. pH had been corrected twice that week. EC was within range at the morning check. Then the lid came off the DWC bucket for a root inspection and the root tips were tan-brown at the base. Not rot, no smell, just that off-colour that means the roots have been running low on oxygen long enough to show it. Water temperature probe went in: 30.8 degrees Celsius. The reservoir was sitting on concrete. It had been sitting on concrete through every previous cycle. The floor was absorbing daytime heat from the room and the building structure and conducting it upward into the bucket base, and we had never measured it because nothing had looked wrong yet.

That one reading connected four things we had been treating as separate problems: the pH that kept climbing faster each week, the EC that spiked between top-offs, the root browning that appeared in week two of hot-season cycles, and the germination inconsistency that only showed up in March. They were all the same problem in different expressions. Bangkok heat loads a hydroponic system in ways that compound on each other, and sensor logs are the only way to see the connections before a crop is already failing.

What the logs showed across six measurement areas

Across 11 logged passive DWC cycles in 2024, mean daily pH rise rate by reservoir temperature band was: below 24 degrees Celsius, 0.06 units per day; 24 to 27 degrees Celsius, 0.08 units per day; 27 to 30 degrees Celsius, 0.10 units per day; above 30 degrees Celsius, 0.14 units per day. Post-acid-correction pH rebound to pre-correction level took 36 to 48 hours at reservoir temperatures above 29 degrees Celsius in passive systems without aeration. Plant Mood grow log data, 2024.

pH creep rate by reservoir temperature band, passive DWC

If you know your reservoir temperature, you can estimate your pH rise rate before measuring. A reservoir at 29 degrees Celsius in a passive system will move approximately 0.10 pH units per day. That means a correction to 6.2 today is a pH of 6.9 in seven days without follow-up. The correction interval writes itself once you have the temperature-to-rate relationship logged for your specific setup.

Cycles where ambient temperature averaged below 29 degrees Celsius showed EC rising from a mean start of 1.38 mS/cm to 1.79 mS/cm at day 7, a 30% increase. Cycles where ambient temperature averaged above 31 degrees Celsius showed EC rising from 1.42 mS/cm to 2.41 mS/cm at day 7, a 70% increase. Same nutrient formula, same reservoir volume, same crop. Daily EC rise correlated with ambient temperature at r = 0.81. Root uptake accounted for an estimated 0.5 litres of daily water loss per day at mid-cycle; evaporation accounted for 1.2 litres per day at 31 to 33 degrees Celsius. Plant Mood grow log data, 2024.

EC trajectory across 7 days by ambient temperature band, NFT tower

A grower who establishes an EC management routine in December and carries it unchanged into April is running a different system than the one they calibrated against. The evaporation rate that drove the October numbers is roughly half the April rate.

Across 8 matched comparison cycles, concrete-floor reservoirs showed a mean daily high of 31.4 degrees Celsius and mean daily low of 27.8 degrees Celsius. Reservoirs on a 5cm timber riser in the same room showed a mean daily high of 28.1 degrees Celsius and mean daily low of 25.3 degrees Celsius. Root browning at the base tips appeared in logged cycles at a mean of 2.4 days before water temperature sustained a 24-hour average above 29 degrees Celsius. The roots signalled oxygen stress before the temperature reading looked alarming on a weekly check schedule. Plant Mood grow log data, 2024.

Daily high and low water temperature by reservoir position across a full season

Managing reservoir temperature reactively, checking only when the plants look off, puts you 2 to 3 days behind the physiological event. The visual signal is not the leading indicator. The temperature log is.

Total daily reservoir volume loss averaged 4.1% of reservoir capacity per day during Bangkok dry season (February to April) versus 2.3% during wet season (September to October) across three NFT systems with 40 to 60-litre reservoirs. Overnight level drop (10pm to 6am, plant uptake near zero) averaged 0.9% of reservoir volume during dry season versus 0.4% during wet season. Ambient relative humidity in the grow space dropped from 74% in September to 54% in March. Plant Mood grow log data, 2024.

Seasonal evaporation rate split, dry versus wet season

The overnight level drop window is clean enough to isolate evaporation from uptake with reasonable confidence. A grower who logs level at 10pm and 6am has the data to distinguish the two loss mechanisms without additional sensors.

Averaged across six tray cycles per EC level at Bangkok ambient conditions: EC 0.8, 93% germination, hull adhesion below 4%; EC 1.2, 91%, hull adhesion 7%; EC 1.6, 74%, hull adhesion 23%; EC 2.0, 51% with cycle-to-cycle variance from 44% to 61%, hull adhesion high. March dry-season cycles at EC 1.6 showed 9 to 11 percentage points lower germination than September wet-season cycles at the same EC with the same seed batch. Plant Mood grow log data, 2024.

Sunflower germination rate and hull adhesion across four EC levels

The EC tolerance ceiling for sunflower germination is not a fixed number. It moves with ambient humidity. A protocol that held at EC 1.6 in October may not hold in March. The guides give a number. The number changes with the season.

Across 12 tower cycles logged in 2024, mean daily pH rise rate by week: week one, 0.08 units per day; week two, drift of 0.5 units over 3 days; week three, 0.6 units over 48 hours; week four, 0.31 units per day. Towers at 80% canopy capacity drifted pH 1.4 times faster than towers at 40% capacity at the same growth stage. Cycles with the fastest pH drift rate in weeks three and four also produced the highest fresh weight at harvest. Root zone pH measured inside root columns ran 0.3 to 0.5 units above simultaneous reservoir readings by week three. Plant Mood grow log data, 2024.

Root-driven pH drift rate by plant load stage, hydroponic tower

A single pH probe at the reservoir wall reads the best-case pH in the system. The roots experience something different. Twice-weekly correction calibrated in week one will under-correct by week four without schedule adjustment.

Plant Mood grow log data, 2024

The numbers that change what you do

pH in a passive DWC system in Bangkok heat moved from 6.1 to 7.6 in 14 days at 0.11 units per day

Across 11 logged passive DWC cycles in 2024 at ambient temperatures above 31 degrees Celsius, with acceleration to 0.16 units per day from day 8 as root mass increased. Iron availability drops sharply above pH 6.5 and is negligible above 7.2. A system spending the second half of a 14-day cycle above 7.0 is running in iron and manganese deficiency for days 9 through 14. Plant Mood grow log data, 2024.

40% faster pH creep rate at reservoir temperature above 28 degrees Celsius versus 22 to 24 degrees Celsius

Comparing matched passive DWC cycles logged at different reservoir temperature bands, pH creep rate differed by 0.04 units per day on average between the two temperature bands, a 40% rate increase attributable to temperature-driven CO2 solubility reduction. Plant Mood grow log data, 2024.

EC rose 64% in 7 days in a Bangkok tower with no nutrient additions; plain-water top-up dropped it to 0.9 mS/cm

Across 9 monitored tower cycles in 2024, reservoir EC averaged 1.41 mS/cm at day 0 and 2.31 mS/cm at day 7. A full plain-water top-up averaged 0.91 mS/cm post-correction, below the 1.2 mS/cm lower bound for leafy green production. Plant Mood grow log data, 2024.

Concrete floor reservoir peaked 3.3 degrees Celsius higher than a raised equivalent in the same room

Across 8 matched comparison cycles, mean daily peak water temperature was 31.4 degrees Celsius on concrete versus 28.1 degrees Celsius on a 5cm timber riser. At 31.4 degrees Celsius, dissolved oxygen saturation in water is approximately 7.35 mg/L versus 8.26 mg/L at 25 degrees Celsius, a 23% reduction in maximum available DO before root consumption is factored in. Plant Mood grow log data, 2024.

Upper tower channel EC averaged 0.41 mS/cm above the simultaneous reservoir reading at 30-minute pump gaps

At 30 to 33 degrees Celsius with 30-minute pump cycle gaps, solution sitting in the upper channels concentrates between pump cycles. If the reservoir reads 1.8 mS/cm, the upper roots may be experiencing 2.2 mS/cm -- enough to produce tip burn while the reservoir reading looks borderline acceptable. Plant Mood grow log data, 2024.

The things that look safe and are not

Across our logged cycles, visual plant health consistently lagged the sensor signal by 2 to 5 days. pH spent 6 to 8 days above 7.0 before interveinal chlorosis appeared in the new growth. Root browning from oxygen stress was visible 2.4 days before water temperature crossed its threshold as a sustained daily average. EC sat at 2.3 mS/cm for days while the plants looked green and growing. The plant signal is real but it is not early. By the time it arrives, the chemical environment that caused it has already been running long enough to reduce yield, even if the symptoms are caught and corrected the same day they appear.

The second counterintuitive finding is that fast pH drift in a well-loaded tower is not a problem -- it is a productivity signal. The cycles with the fastest pH drift in weeks three and four also produced the highest fresh weight at harvest. High root metabolic activity drives both faster pH shift and faster nutrient uptake. A cycle where pH barely drifted through week four was worth investigating for root health problems, not celebrating as a stable system.

The floor placement assumption is a specific version of this error. The instinct to put a reservoir on the floor comes from the observation that warm air rises and the floor is the coldest available surface. In a Bangkok concrete apartment building, the floor surface temperature at 14:00 averaged 30.1 degrees Celsius against ambient air at floor level of 29.2 degrees Celsius. The floor was warmer than the air next to the reservoir. The grower logic that produced floor placement was creating the opposite effect from what was intended. Plant Mood grow log data, 2024.

The standard assumption connecting all four water management variables is that visible plant health is the leading indicator. If the plants look fine, the water is probably fine. pH and EC checks are confirmation. Root browning means something has already gone wrong.

Why Bangkok heat compounds across every water variable simultaneously

The four water management variables -- pH, EC, reservoir temperature, and dissolved oxygen -- are not independent. They share a common driver in Bangkok conditions: elevated ambient temperature acting on a water body with limited circulation, buffer, and thermal insulation. Understanding how they interact explains why managing one variable in isolation frequently makes another worse.

pH creep in a warm passive reservoir is driven primarily by CO2 outgassing. As reservoir water warms above 25 degrees Celsius, dissolved CO2 escapes from solution, taking carbonic acid with it and raising pH. In an active recirculating system, pump turbulence promotes CO2 exchange in both directions, moderating the effect. In a still passive reservoir, outgassing runs unchecked. Adding an acid correction to a passive system above 30 degrees Celsius lowers pH immediately, but the carbonate buffer that would hold the corrected pH has already been depleted by the same outgassing process. The correction is real and temporary. Rebound in 36 to 48 hours is not equipment failure -- it is the physics of a warm, unbuffered water body.

EC drift runs on a parallel but separate mechanism: evaporation. Evaporation removes pure water and leaves dissolved salts behind. In a Bangkok tower at 31 to 33 degrees Celsius ambient, reservoir volume loss to evaporation averages 1.2 litres per day from a 15-litre reservoir -- more than double the plant uptake rate at mid-cycle. A plain-water top-up at the end of a week of drift dilutes the concentrated solution in one step, which overshoots the correction target reliably. The EC went too high because top-off frequency was insufficient for the evaporation rate. The undershoot after correction happens because the correction volume was sized for accumulated drift, not daily maintenance.

Reservoir temperature connects pH and dissolved oxygen. A concrete floor conducts heat into the reservoir base continuously. Water at the base of a reservoir on concrete is measurably warmer than water at the surface. Root tips extending downward sit in the warmest, most oxygen-depleted layer. Dissolved oxygen saturation drops with every degree of warming. At 31.4 degrees Celsius, maximum DO is 7.35 mg/L. In a passive unaerated system with dense week-two root mass, actual root-zone DO is below saturation due to continuous root consumption, meaning effective DO at the root tips during the hottest 4 hours of the day may be closer to 3 to 4 mg/L. That is below the level at which roots function optimally, and it explains the browning that appears in cycles where pH and EC readings both look acceptable.

The root load pH mechanism adds a biological layer on top of all three physical processes. Roots exchange ions with the solution continuously. Most vegetative crops consume more cations than anions at peak growth, releasing hydrogen ions that would push pH down -- but they also release CO2 through respiration into a warm reservoir where CO2 solubility is already reduced, which shifts carbonate equilibrium toward bicarbonate and pushes pH up. The net direction in a typical Bangkok tower at week four is upward. A drift rate that quadrupled between week one and week four in our logged cycles is the expected outcome of a root mass scaling up its metabolic activity in a warm reservoir. It is not a malfunction.

A concrete example from the July 2024 grow logs illustrates the cascade. Ambient temperature averaging 32 degrees Celsius. Reservoir on concrete floor reaching 30 degrees Celsius by day four. pH moved from 6.1 on day 1 to 7.1 on day 8 (rate 0.13 to 0.15 per day). EC climbed from 1.4 to 2.1 over the same period with no nutrient additions. A plain-water top-up on day eight dropped EC to 0.85 mS/cm. pH was corrected to 6.2 the same evening and read 6.9 by the next morning. The plant experienced one day of high EC, one day of low EC, and two days of above-7.0 pH as simultaneous stressors across a 48-hour window, not because of any single error but because three separate physical processes converged on the same correction event. Plant Mood grow log data, 2024.

The ammonium-to-nitrate ratio in the nutrient formula has a direct effect on pH drift direction and speed. Ammonium (NH4+) uptake releases hydrogen ions and pushes pH down. Nitrate (NO3-) uptake releases hydroxyl ions and pushes pH up. Most standard hydroponic formulas are nitrate-dominant, which biases the system toward upward pH drift under plant load. A formula with a higher ammonium fraction can slow drift but carries a risk in Bangkok conditions: ammonium at elevated temperatures promotes algae and can stress roots if it accumulates. It is a real lever but not a cost-free one.

The complaints that are all the same problem

I correct pH to 6.2 and two days later it is back above 7.0. My pH meter must be broken or my pH Down is old.

Neither. In a passive system above 29 degrees Celsius with no aeration, the carbonate buffer that would hold the corrected pH has been depleted by CO2 outgassing. The acid correction works exactly as described. The problem is that in a warm unbuffered reservoir there is nothing left to resist the continued outgassing that raises pH again. A rebound in 36 to 48 hours in Bangkok summer conditions is a physics outcome. Testing with a fresh meter and new pH Down produces identical results. The fix is addressing reservoir temperature and adding surface aeration, not replacing chemistry.

My EC was fine at the morning check and then when I looked again in the afternoon it was way up. I have not added anything.

Evaporation concentrated the solution through the day. EC readings taken at 6am versus 3pm during dry-season days with no top-off showed an average 0.38 mS/cm increase across 18 measurement days in our logged cycles. The mechanism is straightforward: the reservoir loses water to the air continuously, and the dissolved salts stay behind. A morning check captures the overnight equilibration. An afternoon check captures the accumulated midday evaporation. Neither reading is wrong. They are reading a system that changes across the day.

My plants look fine for the first week and then fall apart in week two. I change nothing and the second week is always the problem.

Two things converge in week two. Root mass increases substantially, drawing down dissolved oxygen faster in already warm water. And if the reservoir is on concrete, the floor has been absorbing ambient heat for 7 days and its thermal mass has equilibrated toward room temperature, pushing the reservoir daily low temperature 1.8 degrees Celsius higher on average by day 10 than on day 1. The week-one acceptable readings were the system running inside its tolerance. The week-two failure is the intersection of higher oxygen demand and lower oxygen supply, both driven by the temperature dynamics of a passive system on a warm floor.

My pH was stable for three weeks and then started swinging wildly. Nothing changed.

Root mass crossed a threshold where ion exchange with the reservoir became large enough relative to the buffer capacity of the solution to produce visible drift within a single day. This is normal and predictable. It is not a water quality problem or an equipment problem. The correction frequency calibrated in week one is no longer adequate for the biological load present in week three. The system is working correctly. The maintenance schedule needs updating.

My germination is inconsistent cycle to cycle even with the same seed batch and the same EC setting.

If the germination EC is above 1.4 mS/cm, the inconsistency is seasonal. Lower ambient humidity during Bangkok dry season increases effective osmotic stress on seeds at the same EC level. Cycles run at EC 1.6 during March showed 9 to 11 percentage points lower germination than September cycles with the same seed batch and EC setting. The inputs feel identical because the EC reading is the same. The outcome differs because the osmotic environment around the seed is not.

The advice that does not survive Bangkok conditions

Standard hydroponic guides recommend checking pH and EC twice a week, correcting when readings go outside the target band, and using the same top-off volume across the grow cycle. For passive systems, they recommend placing reservoirs out of direct light to keep them cool.

Each of these recommendations fails in Bangkok conditions in a specific and predictable way. Twice-weekly checks miss the pH acceleration that runs at 0.16 units per day from day 8 in a passive system. A fixed top-off volume calibrated in wet season under-compensates in dry season by 31% on average. And placing the reservoir out of direct light on a concrete floor solves the light problem while creating a heat conduction problem that is larger: the floor at 14:00 in a Bangkok apartment averages 30.1 degrees Celsius surface temperature, conducting heat continuously into the reservoir base regardless of whether light reaches the water surface.

The twice-weekly check failure is the most costly. At 0.11 units per day mean drift rate in a passive system, a reservoir corrected on Monday at 6.2 will read 7.0 by Sunday without an intervening check. That is 6 days of the week spent above the iron availability threshold. Interveinal chlorosis onset in our logged cycles appeared at a mean cumulative exposure of 3.1 days above pH 7.0. A twice-weekly correction schedule structurally guarantees the chlorosis onset threshold is crossed in every cycle. Plant Mood grow log data, 2024.

What the data suggests is coming

Bangkok growers running passive DWC systems without temperature logging will see a measurable increase in unexplained crop failure rates as the city average ambient temperature continues to rise, without identifying reservoir temperature as the cause.

The relationship between reservoir temperature and pH creep rate is log-linear in our data: the rate difference between 27 degrees Celsius and 31 degrees Celsius is larger than the difference between 23 and 27. If Bangkok ambient temperatures increase by even 1 to 2 degrees Celsius as a seasonal baseline over the next decade, passive systems that currently sit just inside the viable range will cross into the range where daily DO deficit is unavoidable without active cooling or aeration. The failure will look like nutrient problems or disease because those are the first things growers check. Without water temperature logs that go back across multiple years, the slow baseline shift is invisible.

A grower who logs daily reservoir temperature and crop failure rate across 3 or more years of Bangkok operation will see failure rate increase in proportion to seasonal ambient temperature change, with the gradient steepest in cycles where reservoir temperature averages above 30 degrees Celsius.

Seasonal EC calibration, meaning adjusting top-off volume and correction frequency at defined seasonal transition points, will become a documented standard practice in Southeast Asian hydroponic communities within 3 to 5 years.

The gap between wet-season and dry-season evaporation rates in Bangkok is large enough that a fixed top-off routine is structurally wrong for approximately four months of the year. This is not a minor edge case. Any grower who logs level and EC at sufficient resolution will see it within one full annual cycle. As logging tools become cheaper and more accessible, the pattern will become visible to more growers simultaneously, and the correction -- seasonal calibration of top-off volume -- is simple enough that it will spread quickly once articulated clearly.

A search of Thai-language and English-language hydroponic growing communities in Bangkok in 2028 and 2029 will show whether seasonal EC management is mentioned in setup guides and grower forums at a meaningfully higher rate than it appears in 2024 to 2026 content.

Root zone pH stratification, meaning the divergence between reservoir probe pH and the pH the roots actually experience inside a dense root column, will be identified as a primary undiagnosed cause of micronutrient deficiency in mature tower systems across tropical urban growing operations.

The 0.3 to 0.5 unit gap between reservoir pH and root zone pH in our week-three and week-four tower cycles is large enough to move iron and manganese availability across the threshold from adequate to deficient. At a reservoir reading of 6.5, root zone pH of 6.9 to 7.0 is already in the lockout range. Growers whose systems are producing interveinal chlorosis in week four while the reservoir probe reads a clean 6.5 are not making a dosing error -- they are experiencing stratification that a single wall-mounted probe cannot detect. As more growers use digital pH probes with logging capability, they will start cross-checking probe placement and discover the gap.

A grower who logs simultaneous pH readings from a reservoir wall probe and from inside the root mass in week three and four of tower cycles across 5 or more cycles will see a consistent gap above 0.2 units in systems with 20 or more mature plants. If the gap is consistently below 0.1 units, this prediction does not hold for well-circulated systems.

One thing to do this week

Pick one logging gap from the list below and close it this week, not next cycle. Not all of them -- one. If you have never logged reservoir temperature at 14:00 in addition to your morning check, do that for three days and compare the two readings. If you have never measured the overnight level drop in your reservoir with lights off, mark the reservoir level tonight and check it at 6am. If you have never taken a pH reading from inside the root mass versus from the reservoir wall in week three, do it at the next inspection. One measurement taken in a place you have never looked before will tell you more about your system than a month of reading about someone else's. The data from our logged cycles is useful as a baseline. Your own logged data is the only thing that tells you what your specific setup does in your specific room in the specific month you are growing. Start there.