Evaporation ponds, denitrifying bioreactors, and reverse osmosis are reshaping how agriculture handles salty, nitrate‑laden drainage — from cheap land-based disposal to 98% water recovery at about $0.46/m³.
By 2030, global freshwater demand is projected to exceed supply by roughly 40% (www.weforum.org). Agriculture — a heavy water user — is under pressure to turn its own drainage into a resource rather than a liability.
The catch: agricultural drainage can be extremely saline and nutrient rich. Tile drains in arid basins such as California arrive at 6–70 dS/m (deciSiemens per meter, a salinity unit; roughly 3,800–44,600 mg/L TDS, or total dissolved solids) and carry millions of tonnes of salt annually (≈0.88 Mt/yr from 39×10^6 m³ of drainage) (www.fao.org). Nitrate-N (nitrate as nitrogen) can exceed 50 mg/L (link.springer.com), and selenium (Se) can reach up to 600 µg/L (www.fao.org).
Those concentrations threaten downstream quality, run afoul of reuse standards, and complicate disposal. The shortlist of options — evaporation ponds, denitrifying bioreactors, and membrane desalination — looks very different depending on the contaminants and end use.
Drainage contaminants and treatment drivers
In many basins, all incoming salts in drainage accumulate in receiving systems; in California’s example, ponds captured ≈22.6 kg of salt per m³ of inflow (www.fao.org). Nitrate-N above WHO limits and trace elements such as selenium add toxicity risks. Indonesia’s water pollution rules require discharges meet strict nutrient/salinity limits, and reuse standards for irrigation or potable use tighten the screws. With scarcity rising, treatment paths must be matched to the dominant contaminant and the reuse or disposal target.
Evaporation ponds (salinity disposal)
Evaporation — or solar — ponds are simple: collect drainage in basins and let the sun do the work. In California’s San Joaquin Valley, 28 ponds covering ~2,800 ha took in 39×10^6 m³/yr of drainage from 22,700 ha of irrigated land, capturing ~0.88×10^6 t/yr of dissolved salts (mostly NaCl, thenardite, etc.) (www.fao.org).
Design notes: in practice, ponds must sit in arid climates where annual evaporation exceeds precipitation; typical layouts keep ≥0.6 m water depth to avoid salt crusting (www.fao.org). They require large footprints — about 12% of the drainage basin in that case — but little power. All incoming solutes remain in the residual brine; volatile contaminants like selenium may partially volatilize or precipitate, with studies showing ~60% of influent Se remains in pond water and the rest immobilized (www.fao.org), yet “sharp bioaccumulation” of Se in birds has occurred even when inflow was ≤2 µg/L (www.fao.org, www.fao.org).
Costs hinge on land and earthworks. Where 1 ha evaporates ~14,000 m³/yr and land is ~$20K/ha, the implied cost is about ~$1.4/m³; broad estimates run ~$20–50K/ha for construction, with very low O&M (vegetation control) (www.fao.org). In some arid basins, evaporation ponds remain “the only economic means of disposal” when no other outlet exists (www.fao.org). They do not remove nitrates or organics; these simply concentrate.
Denitrifying bioreactors (nitrate treatment)
Denitrifying bioreactors — often passive trenches of woodchips — route drainage through carbon media under anaerobic conditions so microbes convert nitrate (NO₃⁻) to nitrogen gas (N₂). Removal rates typically range 1.5–5.4 g N/m³/day; annual load reductions are ~6–55%, with a mean ~30% (link.springer.com). Under favorable design and hydraulic retention time (HRT; the time water stays in the reactor) of ≥6–8 hours, systems reach ~50–70% removal (link.springer.com).
Translating rates: at ~3 g/m³·day and 10‑day HRT, about 30 mg/L N can be removed. In practice, bioreactors often cut concentrations from ~20–50 mg/L to ~10–25 mg/L, while total phosphorus is usually not removed (often initially leached, then sequestered later) (link.springer.com). For projects centered on nutrient targets, suppliers also package dedicated nutrient‑removal systems.
Costs are modest: a U.S. survey (Illinois) found average construction costs ≈US$12,250 per unit (≈$16,000 in 2023 dollars), equating to ~$132/ha‑year treated; woodchip media ran ~$108/m³ (www.sciencedirect.com). O&M is low (periodic flow control; media replacement at ~10–15 years). Median cost per kg N removed was ~$33/kg‑N across monitored sites; “for perspective, removing 5 kg N generally landed at ~$160–$250 per kg (accounting annualized cost), so ~$0.6–1.0 per m³ of water treated (for 20 mg/L inflow)” (www.sciencedirect.com). Government conservation programs can offset ~70% of installation costs (www.sciencedirect.com).
Reverse osmosis (membrane desalination)
Membrane desalination — notably reverse osmosis (RO) — strips out dissolved salts, nitrate, metals, and many organics. A recent conceptual design treating 300,000 m³/d of moderately saline drainage used two RO stages plus thermal/sun ponds to hit ~98% total recovery (≈294,000 m³/d permeate), with the remaining 2% as highly concentrated brine (www.mdpi.com).
The multi‑stage train recovered ~90% in stage one and ~60% of stage‑one concentrate in stage two; laboratory RO alone often recovers 75–90%, while coupling with thermal or solar ponds can push to 95–99% (www.mdpi.com, www.mdpi.com). For plants targeting brackish feeds (maximum TDS around the brackish range), vendors field packaged brackish‑water RO and full membrane systems for large installations.
Costs are substantial. For the 300,000 m³/d concept, capital was ~US$116.4 million (~$396 per m³/day capacity); annual O&M was ~US$48 million, with electricity and chemicals ~75% of O&M, yielding a unit cost of ~$0.46 per m³ (www.mdpi.com, www.mdpi.com, www.mdpi.com). Energy use in that case was ~1.5×10^8 kWh/yr (roughly 5 kWh/m³); at $0.07/kWh that’s ~$0.35/m³ for electricity alone (www.mdpi.com). Literature benchmarks line up: brackish RO typically runs ~$0.30–0.70/m³ for large plants, while seawater RO is ~$0.7–1.4/m³ (www.mdpi.com).
Brine must be managed. In the near‑ZLD (zero liquid discharge) setup above, ~245,000 t/yr of salts (mainly NaCl, MgSO₄, CaCl₂) were crystallized and could be sold at ~$50/t (≈$12.3M/yr) (www.mdpi.com). Without ZLD, the ~2% brine (~6,000 m³/d) would require discharge or disposal (e.g., evaporation ponds). RO requires skilled operation and pretreatment to prevent fouling; in water treatment, pretreatment steps like ultrafiltration are commonly integrated upstream.
Cost–benefit comparison
Different targets, different winners. A few anchors from the literature:
- Land and energy: evaporation ponds use vast land — e.g., 2,800 ha serving 3.97×10^4 ha of cropland (~12%) — yet near‑zero energy (www.fao.org). Bioreactors need <1% of field area and minor power (pumping). RO occupies moderate plant footprints but draws ~3–5 kWh/m³.
- Pollutant removal: evaporation ponds don’t remove nutrients or organics; they concentrate all non‑volatile solutes. Bioreactors selectively remove nitrate (≈30–50% of N load across sites; ≈50–70% under design HRT) (link.springer.com). RO knocks out ~95–99% of TDS, nitrates, metals, and many organics. (Wetlands can deliver moderate N removal of ~40–90% depending on design, but are not detailed here.)
- Water recovery: evaporation outputs 0% water (all evaporated). Bioreactors pass through essentially ~100% of water (minus N₂ gas). RO delivers ≥90% as permeate; hybrids can reach ~98% recovery (www.mdpi.com).
- Capital: evaporation often totals ~$10–50K/ha (land + earthworks). Bioreactors are ~$10–20K each (≈$12.3K average), ≈$132/ha‑year treated (www.sciencedirect.com). RO is roughly $400–500 per m³/day of capacity; e.g., $116M for 300,000 m³/d (www.mdpi.com, www.mdpi.com).
- Operating: evaporation has minimal O&M. Bioreactors are low‑cost to run (media replacement at ~10–15 years). RO runs ~$0.3–0.5/m³, driven by energy and chemicals (www.mdpi.com).
- Unit pollutant removal: bioreactors remove N at ~$25–70 per kg‑N; median ~$33/kg‑N. On a water basis at 20 mg/L inflow, that’s ~$0.6–1.0/m³ (www.sciencedirect.com). RO’s cost per kg of salt removed can be low due to water value; e.g., at ~$0.46/m³ producing permeate from 10 g/L TDS feed, the implied salt removal runs roughly $0.0027/kg.
- Benefits: bioreactors deliver environmental gains by cutting fertilizer runoff; reuse systems in Arkansas were estimated to reduce N outflow by ~50% and save ~21% of irrigation withdrawals, yielding net positive benefit over 40 years (pubs.acs.org). RO enables high‑quality reuse (≈294,000 m³/d in the cited case) and potential salt byproduct revenue. Evaporation ponds are the simplest disposal pathway when land is cheap and climate arid (www.fao.org).
Regulatory context and reuse goals
Indonesia’s national wastewater standards (Permen LHK 5/2014) and irrigation reuse guidance generally target nitrate‑N <10–30 mg/L and salinity low enough to prevent crop damage or groundwater pollution. With untreated drainage often ≈50+ mg/L NO₃‑N (link.springer.com), treatment is necessary for permits or reuse.
For irrigation reuse, saline‑tolerant crops tolerate moderate salinity (e.g., <4 dS/m). In such cases, full RO can be overkill; blending or operational tweaks may suffice. If nitrates are high, a bioreactor can trim nutrient loads before reuse, with much of that nitrogen effectively returned as fertilizer when water is reapplied.
For potable or industrial reuse, only advanced treatment such as RO can meet drinking‑water‑type standards; RO permeate can be blended with fresh supply or directed to specialty irrigation (e.g., greenhouses). For environmental discharge, evaporation ponds plus nutrient uptake in wetlands may meet goals; in the U.S. Midwest, tailwater pools are being adopted to recycle drainage directly into irrigation, capturing ~50% of added fertilizer N on fields (pubs.acs.org).
Design playbook and hybrid trains
Evaporation ponds are most cost‑effective for disposal of high‑salinity drainage when land is available and climate permits, but they do not recover water or remove nutrients; long‑term management of concentrated brines and selenium is a concern (www.fao.org, www.fao.org).
Denitrification bioreactors shine where nitrate is the primary problem, typically removing ~30–50% of incoming N (≈30–70% under design conditions) at about ~$33/kg‑N, with small footprints and low O&M whole‑system analyses report net benefits, including ~50% less N discharge and ~21% freshwater savings on tile‑drained farms (link.springer.com, www.sciencedirect.com, pubs.acs.org).
RO delivers the highest water quality and the largest reuse volumes (≥90% recovery with hybrid trains up to ~98%), but with high CAPEX/OPEX (>$0.4/m³) and brine management duties; it’s best suited to saline‑drainage scenarios where water scarcity is acute or standards are strict, with added upside from byproduct salt recovery (www.mdpi.com).
In many settings, hybridization is decisive: denitrify first (to protect groundwater and downstream waters), then desalinate for reuse, and finally evaporate the small brine fraction. Seasonal storage and reuse can also capture nutrients on‑farm, as shown in Midwestern “drainage recycling” systems (pubs.acs.org). For large membrane builds, operators pair RO with pretreatment and spares; in practice that translates to upstream steps such as ultrafiltration and standardized packages from membrane systems vendors.
Sources: as cited — www.weforum.org; www.fao.org; link.springer.com; www.sciencedirect.com; www.mdpi.com; pubs.acs.org; link.springer.com.
