Arsenic Removal Residential Decision Tree

Why this matters

Arsenic is a Class A human carcinogen. EPA set the federal Maximum Contaminant Level for public water systems at 0.010 mg/L (10 ppb) in 40 CFR 141.62. Private wells are unregulated at the federal level, so the homeowner is the only line of defense. National USGS sampling finds a meaningful share of private wells above 10 ppb, and the distribution is strongly regional rather than uniform: parts of New England, the upper Midwest, and the Southwest run far above the national figure. Look up the county-level data for your service area rather than quoting a national percentage, because the national number understates a bad county badly. Picking the wrong removal technology either fails the lab post-install or costs the homeowner three to five times what they needed to spend. A clean decision tree built on a quality lab result is the difference between a one-time visit and a callback every six months.

Step 1: get a defensible lab result

Before picking technology, the well needs a state-certified lab analysis. A field strip test is screening only and does not distinguish between the two arsenic species.

Required tests:

  • Total arsenic (EPA Method 200.8 ICP-MS, reporting limit 1 ppb or lower)
  • Arsenic speciation (As(III) trivalent vs As(V) pentavalent) by EPA Method 1632 or equivalent
  • pH (As(V) capture varies sharply with pH)
  • Iron and manganese (foul most arsenic media)
  • Silica (competes for sorption sites on activated alumina)
  • Sulfate and chloride (competitors on anion exchange resin)
  • Phosphate (competes on iron-based media and alumina)
  • Total dissolved solids and hardness

Without speciation the design is a guess. As(III) is uncharged at typical groundwater pH (6.5 to 8.5) and passes through every sorptive media that depends on anion charge. The pre-treatment plan must convert As(III) to As(V).

Step 2: pre-oxidation if As(III) is present

Reduced groundwater commonly carries arsenic as As(III). Convert to As(V) before the removal stage:

  • Free chlorine at 1 to 4 mg/L residual with 1 to 2 minutes contact time (fast and reliable; install a small contact tank)
  • Solid-block chlorine pellet feeder or liquid hypochlorite injection at the pressure tank
  • Potassium permanganate at 0.5 to 2 mg/L if iron and manganese are also present
  • Avoid relying on dissolved oxygen alone; conversion is slow and incomplete

Confirm conversion with a post-oxidation speciation test on the next service visit.

Step 3: pick the technology

The three viable residential technologies each fit a defined raw-water window.

Reverse osmosis (point-of-use)

  • Removes both As(III) and As(V) at 90 to 99 percent rejection on a sound NSF/ANSI 58 certified unit
  • Best when arsenic is the only major concern AND drinking and cooking water is the only protected use
  • Wastes 3 to 5 gallons per gallon produced on standard units, 1:1 on permeate-pump or tankless designs
  • Requires pre-sediment, pre-carbon, post-carbon stages and annual membrane sanitization
  • Avoid when bathing exposure or whole-house irrigation contact is a stated concern

Activated alumina (point-of-entry or point-of-use)

  • Effective on As(V) only; mandatory pre-oxidation if As(III) is detected
  • Best at pH 5.5 to 6.0; capacity drops sharply above pH 7.5 and approaches zero by pH 8.5
  • A pH-adjustment stage (CO2 injection or acid feed) is often required ahead of the alumina bed
  • Capacity 2,000 to 10,000 bed volumes depending on raw water; expect annual to biennial media replacement
  • Silica above 30 mg/L and phosphate above 0.5 mg/L sharply reduce run length
  • NSF/ANSI 53 certification for arsenic reduction is the buying spec

Anion exchange (point-of-entry)

  • Effective on As(V) only; mandatory pre-oxidation
  • Works at the natural well pH (no acid feed needed)
  • Sulfate is the primary competitor; sulfate above 25 mg/L collapses arsenic capacity
  • Risk of chromatographic peaking (arsenic discharges above feed concentration as the bed exhausts); use a duplex lead/lag configuration with weekly post-resin sampling, or replace resin on a conservative time schedule
  • Buying spec is NSF/ANSI 61 for the material in contact with the water, plus a device-level arsenic reduction claim under NSF/ANSI 53. NSF/ANSI 44 is the softener standard and covers cation exchange; it is the wrong certification to ask for on an arsenic anion bed
  • Brine waste disposal must comply with local septic and stormwater rules

Iron-based adsorbents (granular ferric hydroxide, iron-modified media)

  • Effective on both As(III) and As(V) at pH 6.5 to 8.0
  • Useful when sulfate or silica rule out alumina or anion exchange
  • Capacity 5,000 to 30,000 bed volumes; single-use media (no regeneration)
  • NSF/ANSI 61 certification for component material

Step 4: the decision tree

  1. Arsenic is the only concern AND only drinking and cooking water need protection: install NSF/ANSI 58 RO at the kitchen sink. Done.
  2. Whole-house protection required AND pH is below 7.5 AND silica below 30 mg/L AND phosphate below 0.5 mg/L: activated alumina with pre-oxidation, optional pH adjustment.
  3. Whole-house required AND sulfate below 25 mg/L AND no other arsenic competitors: anion exchange duplex with pre-oxidation and lead/lag sampling.
  4. Whole-house required AND high sulfate or high silica: iron-based adsorbent with pre-oxidation.
  5. Iron above 0.3 mg/L OR manganese above 0.05 mg/L: install iron and manganese pretreatment first; arsenic media will foul within weeks otherwise.

Step 5: post-install verification

  • Sample post-treatment within 30 days of startup; require less than 5 ppb as the operational target (half the MCL)
  • Sample again at 6 months and annually thereafter
  • For anion exchange: weekly post-resin field sampling during the first six months to catch chromatographic peaking
  • Track bed volumes treated against the design capacity; schedule media change at 75 percent of design capacity

Never recommend a treatment system based on a field strip test alone. The species split changes the technology choice entirely, and customers who relied on a strip-test diagnosis have shipped systems that discharged 40 ppb arsenic from day one.

References

  • EPA 40 CFR 141.62, Maximum Contaminant Levels for Inorganic Contaminants (arsenic 0.010 mg/L)
  • EPA 815-R-05-006, Treatment Technologies for Arsenic Removal
  • EPA Method 200.8, Determination of Trace Elements in Waters and Wastes by Inductively Coupled Plasma Mass Spectrometry
  • NSF/ANSI 58, Reverse Osmosis Drinking Water Treatment Systems (arsenic reduction claim)
  • NSF/ANSI 53, Drinking Water Treatment Units, Health Effects (arsenic reduction claim)
  • USGS Circular 1332, Arsenic in Groundwater of the United States
  • WQA Technical Application Bulletin, Arsenic