Eight samples — tap and well water, each run through four filter setups — tested for metals, salts, and bacteria
against Thai DOH & WHO drinking-water limits. Every sample was visually clear and colourless. The numbers tell a very different story.
LaboratorySuratthani Rajabhat University
Report No.053/68
Sampled / Reported12 Jun 2025
Samples💧 4 tap · 🪨 4 well
Critical finding
Well water carries ~30× the safe lead limit
Raw well water tested at 0.27–0.30 mg/L lead — up to 30 times the 0.01 mg/L drinking limit.
The whole-house Whole-house filter did not reduce it. The point-of-use Point-of-use alone cut it
roughly in half — still 12× over. Only Combined together removed lead completely (Not Detected).
01
Reading the samples
Two water sources, four setups each. The two units tested were a whole-house filter
(point-of-entry, all taps) and an under-sink point-of-use filter (drinking water only).
💧 Source A · Tap Water
Municipal supply
A1No Filterbaseline
A2Whole-housewhole-house only
A3Point-of-usepoint-of-use only
A4Combinedboth combined
🪨 Source B · Well Water
Groundwater
B1No Filterbaseline
B2Whole-housewhole-house only
B3Point-of-usepoint-of-use only
B4Combinedboth combined
02
The safe limits
Drinking-water guideline values (Thai Dept. of Health / WHO). Results are flagged against these.
03
The lead problem in well water
Lead (Pb) measured across all four well-water setups. The dashed line is the 0.01 mg/L safe limit — so low it
barely clears the floor of this chart. Numbers above each bar show how many times over the limit.
Lead concentration · Source B (well)
mg/L · limit 0.01 mg/L
limit 0.01
04
What actually cleans each source
Tap water starts nearly clean; well water does not. Here's the verdict for each setup against the safe limits.
💧 Source A · Tap
Already low in metals. The only flag is trace bacteria.
💧 No filter — metals fine, but total coliform sits right at the 2.2 limit.
💧 Whole-house — coliform rose to 3.6 (over limit); whole-house alone isn't a drinking filter.
💧 Point-of-use — all parameters pass; bacteria not detected.
💧 Combined — all pass; balanced pH 7.07.
🪨 Source B · Well
Heavy lead contamination. Only one setup makes it safe to drink.
🪨 No filter — lead 0.27 mg/L = 27× limit. Unsafe.
🪨 Whole-house — lead 0.30 mg/L = ~30×. No lead removal.
🪨 Point-of-use — lead 0.13 mg/L = 12.5×. Halved, still unsafe.
🪨 Combined — lead Not Detected. Every parameter safe.
05
Every result, every sample
Full lab data. Green = within limit, amber = at/near limit, red = over limit. ND = not detected.
06
What it means
01
Clear ≠ clean
All 8 samples were clear and colourless, yet well water hid lead at ~30× the safe limit. You cannot see it.
02
Whole-house alone won't do it
Whole-house protects the home but left well-water lead untouched. It is not a substitute for a drinking-point filter.
03
Combined removes the lead
Only the combined setup brought lead to Not Detected and passed every parameter — for both tap and well water.
04
Tap needs the point-of-use for bacteria
Tap metals were fine, but trace coliform appeared until the point-of-use filter was in line. For drinking, the under-sink stage is the deciding one.
Part Two
So what should you actually install?
07
Which source should the house use?
If both tap and well are available, the choice comes down to one number — lead. Here's the raw, unfiltered
water side by side.
Parameter
💧 Tap (raw)
🪨 Well (raw)
Limit
Lead (Pb)
ND
0.2687 27×
≤ 0.01
Dissolved solids
10.0
42.1
≤ 500
Chloride
10.9
20.5
≤ 250
Arsenic
0.0001
0.0003
≤ 0.01
Total coliform
2.2
1.1
≤ 2.2
pH
6.51
6.86
6.5–8.5
Everything except lead is comfortably within limits for both. Lead is the whole decision.
💧 Tap — cleaner, but costs to connect
Lead-free at the source. Only needs a drinking-point polish for trace bacteria.
Requires a paid connection — a one-time cost that buys you a safe failure mode: if a filter ever lapses, your worst case is harmless.
🪨 Well — free, but demanding
27× lead straight from the ground — safe only with full metal removal (RO, or Whole-house + DY together).
A metal-bearing aquifer rarely has lead alone, and it's seasonally variable. If a filter lapses here, the worst case is lead in your glass.
Verdict: connect to tap if you can, and treat at the drinking point. Well is viable — but only with stronger, well-maintained treatment, and ideally a broader raw-well metals panel first.
08
A second lab — and a warning about bottled water
A separate ALS-accredited test (Yoga House, Koh Phangan, Mar 2026) compared shop-bought RO refill water
against the point-of-use filter. Different site and source than above — read it on its own.
Parameter
🛢️ Shop RO bottle
↘ Before POU
✅ After POU
Limit
Total coliform
23.0 21×
<1.1
<1.1
<1.1
pH
6.3 acidic
6.6
7.2 balanced
6.5–8.5
Dissolved solids
38
102
122 minerals added
≤ 500
Lead
ND
0.0008
<0.0005
≤ 0.01
Arsenic
ND
0.006
0.006 unchanged
≤ 0.01
Iron
ND
0.09
0.06
≤ 0.3
Manganese
0.002
0.14
0.01 −93%
≤ 0.3
ALS Report 3532129-1, ISO/IEC 17025 accredited. Metals by ICP-MS; microbiology by Standard Methods 9221.
!
Bottled RO failed on bacteria
The shop bottle had 21× the coliform limit and was acidic — but that's the big-blue-bottle problem (stagnant, reused bottles, no chlorine residual), not a flaw in RO itself.
✓
The point-of-use makes "alive" water
It added minerals (TDS 102→122), balanced pH (6.6→7.2), and stripped 93% of manganese — the opposite of RO's strip-everything approach.
·
The point-of-use filter's blind spot
Arsenic passed straight through (0.006, unchanged). Fine here — under the limit — but this filter does not remove arsenic. A high-arsenic source needs RO.
The real lesson: distrust stored water, not the technology. Fresh home filtration beats bottles that have sat for weeks — whatever the membrane.
09
Point-of-use filter vs reverse osmosis — two philosophies
Both reports point to the same split. It's not "better vs worse" — it's a different idea of what clean water is.
Reverse osmosis (RO)
A 0.0001 µm membrane rejects almost everything.
Metals: removes all — lead, arsenic, the untested ones — to near-zero, regardless of load. Bulletproof.
Minerals: stripped to near-zero → flat, acidic water you must remineralise.
Produces some wastewater; benefits from UV to stop tank regrowth.
Point-of-use (selective)
Adsorptive media + 0.007 µm membrane, tuned to keep the good stuff.
Minerals: retained and balanced, pH lifted, bacteria + many metals removed. No waste, no power.
Heavy metals: a finite adsorptive stage — overwhelmed by the well's 27× lead alone, and blind to arsenic.
Best as a drinking-point polish on already-low-metal water.
10
The setup to run
Recommended drinking stack
RO membrane + UV + magnesium pitcher
For a variable source where you may fall back to well water, a countertop RO + UV unit
(e.g. Philips ADD6910: 0.0001 µm membrane + UV-C LED in the clean tank) plus a magnesium pitcher
covers every weakness — each by a different layer.
🛡️
Metals
Lead, arsenic, and the metals nobody tested — removed to near-zero, load-independent.
→ RO membrane (0.0001 µm)
🦠
Bacteria & regrowth
Sterilises the stored clean-tank water on a cycle, so RO's no-chlorine storage risk is handled.
→ UV-C LED
💧
Acidity & minerals
Adds magnesium back and lifts pH — fixing RO water's flat, acidic character.
→ Magnesium pitcher
🏠 Now — the rental (~6 months)
RO+UV dispenser + Mg pitcher. Countertop, no install, moves with you.
Skip whole-house — no point plumbing a place you'll leave.
Keep the RO cartridge on schedule and use the pitcher fresh (it's the one vessel past the UV).
🏝️ Later — the villa (well + tap)
Test the well broadly first — Cd, Hg, Cr, Fe, Mn, hardness, bacteria — before finalising plumbing.
Plumb a dual-source switchover; add a sized whole-house pre-treatment to protect the RO cartridge on gritty well water.
Same drinking stack carries over — the RO+UV unit just comes with you.