Power plants chase ‘zero liquid discharge’ in a thirsty world — but the energy bill is brutal

Thermal power is a water guzzler. A typical coal unit consumes about 2–3 m³ of freshwater per MWh and bleeds off roughly 10–20% of intake as cooling tower or boiler blowdown (pl-cloneforupdate.mystagingwebsite.com). In one design case, a plant pulled 590 m³/h and discharged 124 m³/h (21%) as high‑TDS (total dissolved solids) effluent (nsenergybusiness.com).

The physics are unforgiving: about 80% of intake disappears as unavoidable evaporation from cooling towers and steam cycles, so even a perfect ZLD only saves around 20% of total intake (nsenergybusiness.com). Still, ZLD — which “aims to remove all dissolved solids on-site and reclaim nearly all water,” eliminating brine discharge (nsenergybusiness.com) — is gaining traction where water is scarce or effluent rules are tight.

Policy momentum is real. India requires new thermal plants (post‑2017) to meet 2.5 m³/MWh and achieve ZLD for wastewater (pl-cloneforupdate.mystagingwebsite.com). In China, 65–84% of coal capacity sits in water‑stressed regions, and the Changxing coal‑fired power plant (Zhejiang) commissioned the first forward‑osmosis ZLD system at 650 m³/d (pubs.acs.org). In the U.S., Colorado River salinity controls forced ZLD in the 1970s, speeding permits versus decades‑long discharge negotiations (pubs.acs.org). An influential 2016 review flagged stricter regulations, higher disposal fees, and rising water value as reviving ZLD interest (pubs.acs.org). Market analysts later projected global ZLD/MLD investments around $0.7–1.7 billion per year by the mid‑2020s (sciencedirect.com).

ZLD process train and pretreatment

Power‑plant ZLD targets high‑salinity streams — notably cooling tower blowdown, boiler blowdown, and flue‑gas desulfurization (FGD) effluent. The standard train has three stages: (1) pretreatment — precipitation/clarification or ion exchange — to strip scale‑forming ions and metals; (2) membrane concentration — often reverse osmosis (RO), sometimes electrodialysis (ED) or forward osmosis (FO) — to remove most of the water; and (3) thermal steps — evaporators and crystallizers — to push the remaining brine to solids (pl-cloneforupdate.mystagingwebsite.com). Pretreatment frequently relies on clarifiers; utilities deploy equipment such as a clarifier to remove suspended solids before membranes.

An Indian case routed blowdown through filtration to a membrane (RO/ED) brine concentrator, then into mechanical evaporators (vapor compression) and finally crystallizers to yield a dry salt cake (pl-cloneforupdate.mystagingwebsite.com). For the membrane stage, plants use integrated packages like membrane systems that include pretreatment and RO; in blowdown service that often means brackish‑water RO. Filtration upstream can include pressure UF; in practice, ultrafiltration is a common pretreatment to protect RO. Ion control often uses ion exchange to target specific species ahead of thermal units. Chemical addition in these steps is typically automated with a dosing pump.

The end goal is “~100% recycled water and solid salts for disposal or sale” — but early ZLD systems based on evaporation‑crystallization required considerable energy and capital (powermag.com; pubs.acs.org). Sanitary or heavy‑organic streams are typically excluded or sent off‑site, per design guidelines (nsenergybusiness.com).

Evaporation ponds: land for energy

Evaporation ponds — large, shallow basins that rely on solar/ambient heat to evaporate water and leave salts behind — are the low‑tech ZLD option (condorchem.com). They “utilize natural solar energy” and carry low operating costs, assuming a high evaporation rate and inexpensive land (pubs.acs.org).

The trade‑offs are severe: they are climate‑limited, need vast rural land, and cannot recycle the evaporated water — it is lost to the atmosphere (pubs.acs.org). In practice, land acquisition alone can be ~3× more expensive than a comparable mechanical ZLD plant (pubs.acs.org). In humid or rainy regions, overflows and odor become serious issues (condorchem.com; condorchem.com). The net: ponds are viable only for relatively low flows in very dry climates.

Brine concentrators (MVR): compact, energy intensive

Mechanical brine concentrators evaporate and condense water using mechanical vapor recompression (MVR) or thermal vapor compression. Falling‑film designs are often seeded with scale crystals (e.g., gypsum) to minimize fouling (powermag.com). Typical MVR energy use runs roughly 18–35 kWh (electrical equivalent) per metric ton (~1 m³) of water evaporated (powermag.com); an ACS review cites ~20–25 kWhₑ/m³ for brine concentrators (pubs.acs.org).

The upside is a small footprint and near‑total capture of evaporated water for reuse; the downsides are high capital and operating costs and careful scale control. Operators typically dose scale control chemistry; programs like a scale inhibitor are standard alongside membrane antiscalants in the upstream RO step. Thermal ZLD with concentrators and crystallizers has been deployed in industry for roughly 40 years, and many U.S. plants still use MVR evaporators for FGD wastewater ZLD (pubs.acs.org; powermag.com).

Crystallizers: the last, costly mile

After a concentrator removes ~75–95% of water, a forced‑circulation crystallizer drives the brine to salts by boiling and crystal formation followed by solid/liquid separation (e.g., filter press). One source notes a crystallizer “requires slightly more than a metric ton of steam to evaporate a metric ton of water” — roughly 600+ kWh thermal per m³ — and the final push often exceeds 30–40 kWhₑ per m³ on an electricity‑equivalent basis (powermag.com). Full ZLD to solids can cost on the order of $8–11 per m³ just for steam/electricity (nsenergybusiness.com). Plants facing less stringent rules sometimes stop at a high‑TDS brine, disposing off‑site or using it for ash wetting — which slashes operating costs but is not strict ZLD (power-eng.com).

Membranes reduce thermal load

Practical ZLD combines technologies, with membrane steps shrinking the thermal duty. RO or ED/FO can cut feed volume to a few percent so only a brine goes to the concentrator and crystallizer (pubs.acs.org). Where climate and land cooperate, evaporation ponds can replace a crystallizer — but at the cost of lost reuse water (pubs.acs.org). For blowdown concentration, plants often standardize on RO packages; catalogs of RO, NF, and UF systems are geared for industrial duty.

A water balance reality check

NS Energy illustrates the ceiling: without ZLD, a plant might draw 590 m³/h raw water and waste 124 m³/h (~21%) as discharge. With ZLD, about 99.5% of that 124 m³/h is recovered, taking the waste down to ~0.1% and saving roughly 118 m³/h — about 20% of fresh needs (nsenergybusiness.com; nsenergybusiness.com). Most water is still lost to evaporation and cannot be recovered (nsenergybusiness.com).

Capital intensity and operating costs

ZLD is expensive. Industry presentations peg ZLD at roughly 1.4–3× the cost of simpler biological FGD treatment at mid‑scale flows (up to ~200 gpm) (power-eng.com). One analysis found a thermal ZLD system (brine concentrator + crystallizer) had ~40% higher installed cost than a biological alternative (power-eng.com; power-eng.com). Levelized costs of recovered water land around $7.8–13.8/m³ in surveys (mdpi.com).

Energy dominates OPEX: electricity/steam can be ~70% of ZLD operating cost (mdpi.com). MVR evaporation runs ~18–35 kWh/m³ (powermag.com), while full crystallization can demand ~1,000+ kWh thermal per m³. In dollar terms, completely evaporating water can cost about $32.82 per 1,000 gallons (≈$8.7/m³) in steam (power-eng.com). A 10,000 m³/d ZLD unit at $8/m³ implies roughly $8 million per year in O&M chemicals (lime, acids) and labor add further costs. In one study, RO+crystallizer produced reclaimed water at ~$7.85/m³ versus ~$2.85/m³ being paid locally for raw water (mdpi.com; mdpi.com).

Beyond equipment, engineering and permitting can add ~10–15% of project costs (samcotech.com). Evaporation ponds need liners and large real estate; mechanical plants require structural foundations and large heat‑exchange surfaces. A turnkey MVR+crystallizer at 10–20 m³/h can run several million dollars. Financing at ~9–15% annual interest is common; viability often hinges on local water and power prices, permits, and offtakes. Support systems — from pumps to controls — are part of the build; many suppliers bundle water‑treatment ancillaries into ZLD packages.

Benefits and risk mitigation

The direct win is water reuse. Recovering ~5,000 m³/day of blowdown (typical for a ~1 GW plant) at a $0.50/m³ tariff yields ~$2,500/day (~$0.9 million/year) in avoided purchases; at $2–3/m³ in some arid regions, the savings rise to several million USD per year (mdpi.com). Disposal savings can also be material: one case reduced brine volume and mixed it with ash, cutting a typical ~$32.8/1,000 gal disposal cost to about $4.4/1,000 gal (power-eng.com).

Strategically, ZLD unlocks compliance and permits that might otherwise be impossible under tight TDS and metals limits, avoiding fines, shutdowns, or relocation (pubs.acs.org). It also improves water‑supply reliability in drought‑prone regions. Byproduct credits can help: desalination‑ZLD studies found sale of recovered salts covering ~25–30% of revenues (mdpi.com). A brine‑valorization analysis projected profits of about $180–226 per day from selling reclaimed water and salts (context: low brine volumes <1 m³/d in generating industries) (sciencedirect.com).

When ZLD pays, and when it does not

On pure economics, many analysts long deemed ZLD “not viable” due to high energy and capital intensity (pubs.acs.org). But mandates flip the calculus. India’s NTPC is equipping new stations for ZLD despite the cost, expecting long‑term savings in makeup water and smoother permitting (pl-cloneforupdate.mystagingwebsite.com).

A quantitative frame: a 500 MW coal plant (capacity factor ~80%) at 2.5 m³/MWh uses ~24,000 m³/day intake, with ~4,800 m³/day blowdown (20%) (pl-cloneforupdate.mystagingwebsite.com). If ZLD recovers 98% (~4,700 m³/d), and raw water is $1/m³, annual savings ≈ $4.7 million. With a $30 million installed ZLD and ~$3 million/year O&M, simple payback (ignoring time value) is about (30M)/(4.7–3.0) ≈ 26 years. If raw water were $3/m³ or disposal savings add ~$2M/year, payback shortens to ~10 years. Studies see 5–10‑year paybacks only in favorable scenarios (high water cost, byproduct credits, waste‑heat use); one techno‑economic model estimated ~7.7 years (mdpi.com). For context beyond power, Beijing’s municipal reuse program reported benefits ≈1.7× project cost (researchgate.net).

Planning metrics for utilities

Useful benchmarks from industry and reviews:

• Water recovered: typically 95–99% of blowdown; one case showed ~99.5% recovery from 124 m³/h blowdown, cutting waste to ~0.1% (nsenergybusiness.com).
• CAPEX: roughly $1,000–5,000 per m³/day capacity (ballpark; higher for very high TDS).
• Energy use: MVR ~20–30 kWhₑ/m³ on average (pubs.acs.org); crystallization adds ~600+ kWh thermal per m³ for the final push (powermag.com).
• OPEX: around $2–10/m³, with 68–70% of OPEX being energy (mdpi.com).
• Payback: typically >5 years in high‑cost scenarios; often 10+ years if water is ~$1–2/m³ (mdpi.com; mdpi.com).

Bottom line for water‑stressed grids

ZLD is technically feasible and increasingly mandated. Where land and climate permit, ponds swap power for acreage; elsewhere, MVR concentrators and crystallizers carry the load — and the energy bill. The decision turns on local prices and permits as much as physics: a comprehensive analysis should weigh CAPEX (in the millions), OPEX ($/m³), water savings (often <$1–3/m³), disposal/penalty avoidance, salt credits, and regulatory risk mitigation (pl-cloneforupdate.mystagingwebsite.com; pubs.acs.org). For many utilities, the real return is license to operate.