Converting a lime softening plant to membrane softening

A regional utility’s 30 ML/day groundwater plant has supplied a mixed industrial and municipal system for decades. Its original lime softening train removed carbonate hardness, but ageing vessels, variable lime feed and unreliable recarbonation made finished-water alkalinity difficult to control. The utility now faces tighter discharge conditions, higher sludge handling costs and demand for steadier water quality.

This case study follows the site from operating history through a hybrid membrane alternative. It shows how engineers can compare calcium carbonate removal, Langelier Saturation Index (LSI) control, membrane recovery, concentrate disposal and operator workload, then translate those findings into a conversion package that can be built, monitored and audited.

A representative Australian conversion case

The assessment begins with a groundwater blend averaging 280 mg/L hardness as calcium carbonate, alkalinity around 210 mg/L and total dissolved solids near 650 mg/L. Seasonal changes raise turbidity and iron, while a new industrial connection increases the need for predictable softened water.

The proposed arrangement retains two concrete equalisation basins, repurposes the clearwell and installs skids of nanofiltration for bulk hardness removal plus reverse osmosis polishing for the highest-quality stream. Existing lime reactors provide standby capacity during commissioning, avoiding an immediate dual-operation requirement.

Success is defined through agreed product hardness, recovery, availability and concentrate criteria. The utility also requires a whole-of-plant cash flow, workforce transition plan and permit pathway, so membrane selection alone cannot be treated as the completed solution.

Chemistry behind the conversion

In lime softening, calcium precipitates as calcium carbonate while magnesium commonly precipitates as magnesium hydroxide. The process raises pH, creates sludge and requires recarbonation to stabilise finished water. Performance depends on lime availability, mixing, temperature, settling and chemical dosing rather than a single control point.

Nanofiltration rejects much of the divalent ions responsible for hardness but retains more monovalent salts than reverse osmosis. Design therefore starts with a full ionic balance, not hardness alone. LSI, calculated as pH minus the saturation pH, indicates carbonate scale tendency, while silica, sulphate, barium and boron can impose separate limits at higher recovery.

Pilot testing held concentrate LSI below the scaling threshold while varying nanofiltration recovery from 75 to 85 per cent and polishing recovery from 70 to 75 per cent. Those results established a defensible operating envelope linked to the worst credible source-water blend and verified through a chemical-cleaning trial.

Membrane selection and pretreatment

The operating envelope supports a modular train rather than a single large membrane bank. Nanofiltration units handle bulk softening and most of the flow; reverse osmosis treats a selected fraction for low-conductivity water. Variable-speed pumps recover energy when demand falls, while the existing clearwell provides buffer storage during cleaning or peak production.

Pretreatment is the main safeguard. Media filtration or dissolved air flotation removes suspended material, and cartridge filters protect the elements. Oxidised iron must be filtered, residual chlorine removed where required, and feed SDI kept below the membrane supplier’s limit. Antiscalant, acid or biocide dosing is selected from pilot chemistry.

Each train needs conductivity, pH, flow, pressure and temperature measurements, together with differential-pressure and normalised permeate-flow trending. Cleaning triggers should be tied to verified membrane conditions rather than operator guesswork, and spare elements should be stored under supplier-specified temperature and humidity controls.

Recovery, concentrate and mass balance

Recovery is the central operating and environmental variable. At 80 per cent overall recovery, a 30 ML/day plant produces about 24 ML/day of usable permeate and 6 ML/day of concentrate. Water recovery is not pollutant removal: rejected salts enter the smaller stream at much higher concentrations, so the disposal route must be established before the recovery rate is fixed.

Australian disposal options include dilution to a sewer with adequate treatment capacity, controlled land application, evaporation, lined treatment systems and, in some locations, managed borewells or zero-liquid-discharge projects. Each carries different consent, monitoring and public-acceptance requirements. Sewer dilution may shift the salt load rather than eliminate it, and recycled-water networks can restrict concentrate where salinity could affect soil or downstream users.

The waste account must include cleaning chemicals, cartridges and membrane replacements. A broader inventory method similar to the one in this household waste footprint guide can help teams separate avoided lime sludge from new concentrate quantities. Reporting both water and contaminant mass prevents higher recovery from appearing artificially beneficial.

Commissioning, operators and process control

Transition begins with a management-of-change review covering isolation, temporary bypasses, laboratory methods, dosing adjustments and permit conditions. During commissioning, operators compare new measurements with the old process, establish stable baselines and record deviations from the pilot envelope. Automatic optimisation remains disabled until the control logic has survived a full operating cycle.

Existing lime operators understand hardness chemistry, settling and safe chemical handling. They also need membrane-specific competency in normalisation, salt passage, cleaning chemistry and fault diagnosis. Automation can control ramp rates, flux limits and cleaning requests, but it should flag conditions for review rather than conceal a failed analyser or improperly fouled train.

Cross-jurisdictional peer learning can be supported through technical committee volunteering, while local training should include site drills and competency sign-offs. Imported membrane elements and pumps need accessible spare parts, agreed freight arrangements and technicians capable of reaching the plant quickly.

Economics, standards and Australian realities

The financial comparison uses Australian dollars and a life-cycle period matching the asset plan. It includes membrane replacement, energy at each tariff period, chemicals, sludge transport, cleanings, labour, instrumentation and the value of recovered capacity. Imported elements expose projects to exchange rates, shipping and customs clearance, so tender evaluation must consider local support commitments as well as headline capital cost.

Local source and disposal conditions differ sharply. Sydney combines dense recycled-water networks with strict source-water expectations in New South Wales. Adelaide’s water-security planning makes salinity and reuse central to investment decisions. Perth’s mineralised groundwater can increase pretreatment and concentrate-management requirements. Brisbane and the wider South East Queensland system must accommodate growth, flood events and variable raw-water quality. These conditions favour robust design and staged investment over direct transfer from one overseas analogue.

Each state’s environmental authority, water utility and health regulator may impose different sampling, consent and reporting duties. Equipment should meet applicable Australian and New Zealand standards, with electrical, pressure-vessel and workplace requirements allocated clearly in the contract. Australian utility-style tenders also benefit from guarantees tied to whole-of-plant performance rather than individual membrane capacity alone.

Stakeholders, performance and delivery

The operator, regulator, industrial customers and ratepayers each need a different view of the conversion. Industrial users may value consistent hardness more than stringent conductivity, while regulators require detailed concentrate and sludge data. The Brisbane Diamantina consumer representative guide demonstrates the value of involving consumer representatives early, even when the end user is primarily industrial.

The decision case projects a substantial reduction in lime sludge, more stable finished-water alkalinity and lower routine chemical handling. Energy use remains sensitive to feed temperature and recovery, so savings are not assumed. Twelve months of operating data should test the projections, including membrane ageing, cleaning frequency, availability and the actual cost of concentrate disposal.

A staged implementation reduces risk: confirm source-water variability, validate pretreatment, commission one membrane train and release the remaining capacity after performance review. The next step is a 30-day site pilot that produces a signed mass-balance and treatment-performance design basis for the conversion.