Ion Exchange: A Viable Water Treatment Alternative To Membranes

By Gary Thorne, Parsons Brinckerhoff and Joel Segal, Melbourne Water

Five decades ago, ion exchange using charged resins was one of two processes used in the water industry for water treatment. The ion exchange (IX) process involves ‘swapping’ an undesirable dissolved salt – for example, arsenic, calcium, or radium – for a more benign ion such as sodium. In the late 1980s, membranes used in reverse osmosis (RO) were becoming the favoured process, but IX never disappeared. IX evolved from a chemical-based regeneration process to one that still used chemicals but also used positive and negative electrical charges to regenerate the resin.

Parsons Brinckerhoff has used electro-deionisation in several power station designs and more recently has started to look at selective resins for treating specific water contaminants in remote communities and acid mine wastes.

Water Treatment In Remote Communities

One of the key advantages IX and adsorption processes have over RO in remote areas is the waste volumes of water produced. For every volume of water treated by RO, anywhere from 10 percent to 50 percent of that amount is rejected as a concentrated brine stream. With IX, only the spent regenerant chemical and some of the rinse water are wasted, which is typically only 2 percent to 4 percent. This preserves much-needed water in remote, arid areas where a consistent water supply is often a challenge.

Since 2005, Parsons Brinckerhoff has managed the Remote Areas Essential Services Program (RAESP) on behalf of the Department of Housing in Western Australia. A recent project involved the treatment of a communal water source that had high nitrate levels. The presence of high silica levels restricted the water recovery ratio of an RO system to only 50 percent to 60 percent, so alternative processes such as nitrate-selective IX resin were considered because of the reduced volume of water wasted and the improved nitrate removal over a six-week period.  The advantages for this project were the following:

• IX waste volume could be sent to an existing sewage evaporation pond, whereas an additional evaporation pond area would be required for the membrane processes (RO and EDR[1]) waste volume (see Table 1);

Table 1 – Waste volumes of various water treatment processes

• In addition to the wastewater volume, once evaporated, the membranes produce twice the dried weight of salts, which also has to be disposed of (see Table 2);
• Reduced power costs (see Table 3);
• Chemical regenerant is a salt solution (NaCl), whereas the membrane (RO and EDR) systems required a specialty antiscalant and other chemicals, such as HCl, to be provided on a regular basis to clean the membranes (see Table 4); and
• Longer resin life compared to membrane life.

The same principle of selective resins was applied to the removal of uranium from a remote Australian community’s water source, and has reliably produced quality drinking water since its installation nearly two years ago.

Table 2 – Dried waste stream of water treatment processes

Table 3 – Power costs of water treatment processes

*Based on the maximum NaCl cost of $250/tonne

Table 4 – Chemical and maintenance costs of water treatment processes

Water Treatment in the Mining Industry

Another field in which Parsons Brinckerhoff has specified IX systems is in the mining industry for acid mine drainage. Two Asian copper mine projects required large volumes of river water to be pumped for use in the mine operations. If standard dam management principles were applied, the extracted water would no longer be available to communities downstream whose water demand was already near capacity in the dry season. An additional issue facing the mines was the extensive wet season, which removes the possibility of using solar evaporation and increases the potential of flooding to waste storage dams.

Treating the waste stream for compliance with the local environmental protection requirements could be easily achieved, but in each case the mine owner recognised that having a sound environmental approach was the only way forward. It was, therefore, imperative that the project did not reduce water flow or quality to downstream users. The Australia and New Zealand Environment and Conservation Council (ANZECC) water quality guidelines were applied to bring the discharge limits back to receiving water quality levels. This requirement promoted the recycling of the tailings dam and drainage water, while stringent discharge water quality targets fostered the need for a high level of water treatment to achieve combined treated water flows of up to 172 ML/day.

An integrated, phased water treatment approach was identified as the most cost-effective treatment strategy. After extensive market evaluation, there were no similar plants operating in Australia, so due diligence audits of available treatment processes (in the US and China) were undertaken. The findings were added to the Parsons Brinckerhoff “bow tie” risk-weighted net present value (NPV) options analysis methodology, which is based on ISO 31000 Risk Management technique B.21 (see Figure 1). The process details the various pathways to potential loss, including environmental damage, and those controls critical to the prevention and mitigation of loss.

Figure 1 - An example of Bow Tie Risk analysis

Two viable outcomes formed a common approach to the pre-treatment of metal precipitation which is followed by the RO or IX process to remove sulfates.

• Reverse osmosis with conventional softening through the metal precipitation process typically achieved a 65% recovery ratio and provided permeate (product water) with a composition that could negatively impact the receiving ecosystem. Remineralisation to restore the hardness, alkalinity, and pH to background levels was a requirement. Due to the humidity and rainy seasons, evaporation ponds were not feasible, and so the RO waste concentrate stream would require further treatment through brine concentration and crystallisation.
• The IX process would regenerate on a 12-hour cycle and provide approximately 90% efficiency. The IX process can efficiently operate on gypsum-saturated waters to achieve final sulfate concentrations of 200 to 500 ppm, while producing a pure gypsum (CaSO₄) by-product. The IX process removes calcium and sulfate from wastewater in order to achieve effluent compliance with sulfate discharge limits, which has the added advantage of reducing final effluent total dissolved salt (TDS) concentration. The cation and anion resins are regenerated using sulfuric acid and lime, respectively, in both cases generating gypsum, which is precipitated using a seed. The gypsum by-product in its solid form can potentially be reused.

The IX waste stream is dewatered using a plate and frame membrane press, eliminating the need for crystallisation. The press filtrate is returned to the head of the treatment stream and the pure, solid gypsum can be sold to the building material industry or used in soil augmentation.

When comparing the two processes, the benefits of the IX system are:

• Capital cost is about 40% less, as the concentration/crystallisation stage was not needed;
• Operational costs are about 50% less (mostly due to the greatly reduced power demand); and
• There is approximately 80% less CO₂ emissions due to the much lower power requirements.

This initiative not only transformed a potential environmental threat into a sound environmental opportunity, but also demonstrates the feasibility of large-scale alternatives to current wastewater management strategies.

Developing a Water Treatment Strategy

The current water industry practice is often to focus on membrane technologies, which have taken a large share of the desalination market. Particularly for brackish water, membranes have completely superseded IX as the technology of choice. Today IX is making a comeback for the treatment of natural, surface, or deep well water. Parsons Brinckerhoff is assisting an IX technology company to develop a process to treat coal seam gas water and brackish water, two major treatment challenges currently facing the Australian mining industry. A member of the Parsons Brinckerhoff treatment team developed the tri-bed desalination process with the assistance of resin manufacturer Rohm and Haas, and the first working process was installed and operating in the 1980s.

The introduction of improved weak base anion and weak acid cation resins, which result in lower regeneration costs and higher exchange efficiencies, means the process can be a cost effective alternative to membrane systems on lower salinity brackish water. The original tri-bed process was a three-pass system:

1. Alkalisation - Resin is in the bicarbonate form, which will remove the associated anions (chlorides, sulfates, nitrates, etc.). The final products from the unit are calcium bicarbonate, magnesium bicarbonate, etc.
2. De-alkalisation - Resin is in the hydrogen form. Cations (calcium, magnesium, sodium, etc.) are taken up by the resin, with only water and carbon dioxide leaving the unit.
3. Carbonation - Weak base anion resin in the hydroxyl form takes up the carbon dioxide to convert the resin to the bicarbonate form, with desalinated water leaving the treatment plant.

Regeneration of the first two units is carried out in external regeneration columns with sulfuric acid and lime, while the final stage resin is transferred to the first vessel to be utilised for its bicarbonate form. Testing has shown the process can be adapted to operate efficiently over a range of brackish water salinities (up to around 6000 ppm).

Costing studies have been conducted by various companies to help clients determine under what scenarios IX or RO is the more appropriate and cost-effective choice. However, rather than viewing them as competing technologies, we are seeing a move towards viewing them as technologies with overlapping, but also individual, niche markets as evidenced by the highly selective nature of certain resins. Moreover, they are more commonly used as complementary technologies, where IX reduces the scaling potential on high salinity water or polishes the membrane permeate, to achieve a difficult water quality target with minimal waste streams.

Gary Thorne is a Principal Water & Process Engineer in Parsons Brinckerhoff’s Brisbane, Australia office. He has over 35 years of experience in the consulting and contracting professions in developed and developing countries. His experience includes conceptual and detailed design, construction, commissioning, testing, and operation of a wide range of water, wastewater, and sludge treatment plants.

Joel Segal is a Process Engineer at Melbourne Water currently involved in capital project planning for the 500 ML/d Western Treatment Plant. His experience is in biological wastewater treatment and municipal and industrial desalination. He was a former Graduate Water & Process Engineer at Parsons Brinckerhoff.