With the treatment of flue gas desulphurization wastewater becoming mandatory in coal-fired plants, Prakash Govindan and Zhifeng Fan make the case for carrier gas extraction as a cost-effective and performance-boosting solution

Cost-effective treatment of FGD wastewater is becoming mandatory in all coal plants

Credit: Gradiant

Flue gas desulphurization (FGD) wastewaters are produced at coal-fired power plants in increasing quantities as the regulation on air emissions is tightened worldwide.

Low cost and environmentally favourable reuse of this wastewater stream has become an important topic with the respective national and local regulatory bodies stipulating minimum treatment levels and standards.

Traditional technologies, which are otherwise used for concentration of saline streams, fail the economic and performance benchmark that needs to be met.

Carrier gas extraction (CGE) technology, which was specifically developed to handle high levels of contamination and variability in feed waters, is optimal for this application and offers the lowest cost solution within the required performance.

Source and constituents

Coal-based power plants generate over a third of the planet’s electricity. The combustion of coal in these facilities produces a flue gas that is emitted to the atmosphere.

Many power plants are required to remove SOx emissions from the flue gas using FGD systems. The leading FGD technology used globally is wet scrubbing (85 per cent of the installations in the US and 90 per cent of the installations in China).

Commonly, three kinds of scrubbers are used for wet scrubbing – venturi, packed, and spray scrubbers – and entail injection of alkaline scrubbing agents into the scrubber.

Typically, the agent is limestone (i.e., calcium carbonate), quick lime and caustic soda.

For example, when limestone reacts with SOx in the reducing conditions of the absorber, sulphur dioxide (the major component of SOx) is converted into sulfite, and a slurry rich in calcium sulfite is produced. In forced oxidation FGD systems, an oxidation reactor is used to convert calcium sulfite slurry to calcium sulfate (gypsum).

From the blowdown and dewatering processes of the slurry, the FGD wastewater stream is created. The composition of coal and limestone primarily affects the composition of the wastewater. Other parameters which have a smaller effect are the type of scrubber and the dewatering system used. Coal contributes chlorides, fluorides and sulfate to the wastewater. Because of the metallurgy used in the scrubbers it is typical to purge the wastewater before the chlorides exceed 12,000 mg/L. The level to which chlorides are tolerated by the metallurgy of the scrubber determines the amount of wastewater generated. Use of better metallurgy can help reduce the amount of wastewater by not purging until a chlorides level of up to 35,000 mg/L is reached.

Coal also adds trace metals, including arsenic, mercury, selenium, boron, cadmium and zinc. Limestone could contribute iron and aluminium to the FGD wastewater.

Varying water quality

Defining a standard composition of FGD wastewater across different power plants is tricky because there is no consensus in the industry on where (in the process train) the sample for measuring the composition has to be collected, and the design of the process train downstream of the scrubber itself changes from one facility to another. For example, some facilities may employ primary and secondary hydro cyclones to maximize the capture of solids before gypsum dewatering.

It is also common for a plant to change coal and limestone suppliers so the wastewater constituents will change over time during operation of the FGD system. Unlike other industrial wastewater treatment fields, FGD wastewater samples for a specific plant are likely not available for testing before the plant is designed, built and commissioned. Additionally, during the operation of the coal-fired power plant, there might be periods when the plant is not run at full capacity and the SOx levels and FGD water quality and volumes can vary.

Figure 1. Schematic of carrier gas extraction demonstrating the production of fresh water and heat recovery within the system


Figure 2. Relative condenser sizes. New CGE BC devices have 40 per cent lower surface area requirement than vapour condensers

Complicating matters further, the plant’s wastewater treatment system must be flexible to handle these varying inputs yet produce a treated stream that meets the plant’s wastewater discharge permit requirements. All traditional treatment technologies fail because they cannot handle these requirements as they were developed for sea water or other applications where the feed water quality and volumes remain fairly constant.

CGE technology

Invented at Massachusetts Institute of Technology, CGE is a novel method of desalinating high salinity water streams using a carrier gas. The technique was specifically developed to handle high contamination wastewaters at varying volumetric rates and quality. Over 50 patent families cover various innovations which make CGE an economical technology for treating FGD wastewaters to high recovery rates and with high levels of influent water variability.

CGE is a desalination process that mimics the rain cycle. It uses a carrier gas as a medium to desalinate saline streams by using a humidification-dehumidification configuration. CGE consists of two main unit operations: humidifier and dehumidifier (see Figure 1).

Both the humidifier and the dehumidifier are direct contact heat and mass exchange devices. The humidifier is a packed bed device wherein a heated water stream (<90°C) is introduced in the form of droplets and the carrier gas (which is typically ambient air) is introduced at the bottom of the device in a counter-current configuration.

The carrier gas comes in direct contact with the saline droplets and there is evaporation from the surface of the droplet into the carrier gas stream. Hence, as the carrier gas rises through the device, it accumulates increasing amounts of pure water vapour from the saline stream. The concentrate, which retains all of the dissolved salts, exits the humidifier, is diluted with feed wastewater, is preheated with heat from the dehumidifier, is heated using a source of energy (like solar heaters) and is recirculated back into the humidifier.

The dehumidifier is a multi-stage bubble column device. In this device, the air-vapour mixture from the humidifier is sparged through several shallow layers of fresh water in successively cooler stages. As a result, small bubbles are formed and the vapour condenses from the surface of the bubbles into these shallow pools. As this fresh water is generated it also picks up the heat of condensation, which is transferred back to the feed water in the preheating heat exchanger.

CGE was developed to handle high levels of contamination

Credit: Gradiant


Figure 3. Relative scaling levels. CGE can take orders of magnitude high thickness of scale because of the separation of the heat exchange surface from the phase change surface

The ratio of the energy input in this preheating process to that in the heater that follows it (energy recovery ratio) has been maximized over a decade of research and development. Currently, CGE’s energy recovery ratio is up to 4.5 which implies that only ~22 per cent of the total heat needed to distil the water is provided by the heat source and contributes to the energy cost. Additionally, a relatively low top brine temperature of <90°C combined with a low temperature increase required in the heater (and no steam requirement) makes the CGE optimal for using simple solar thermal energy (like evacuated tube solar heaters) as heat input, or using hot water from power plant heat recovery steam generators.

In addition to treating wastewater at full scale and in the lab to purity levels, the performance of CGE has also been tested with more than 100 different high contamination FGD wastewater samples provided by various industrial customers, including power plants.

The recovery of the system depends on the feed concentration as saturated brine solution is the waste/product stream leaving the system. With feed water streams of less than 120,000 ppm TDS, recovery rates are consistently greater than 95 per cent.

Treatment of industrial wastewaters is limited due to the prohibitive capital investment and operating costs associated with distillation-based techniques like mechanical vapour compression (MVC). These treatment options are expensive because:

1) These technologies were not developed to handle variability in feed water quality;

2) FGD wastewater is corrosive, requiring expensive corrosion-resistant metal for evaporation and condensation surfaces;

3) The conventional treatment technology, i.e., MVC is not energy efficient at high salinity and the corresponding high boiling point elevation (as high as 13°C at saturation concentration of sodium chloride); and

4) Pretreatment requirements are much higher with high hardness levels because the evaporation and condensation surfaces are sacrosanct and cannot accumulate porous scale layers (if they do the performance of the system drops precipitously).

Since its commercialization, CGE has demonstrated a significant shift in the cost of treatment of hypersaline waters based on three main innovations.

Bubble column (BC) heat exchangers (see Figure 1) have extremely high heat and mass transfer rates as they employ direct contact condensation of the vapour-gas mixture in a column of shallow liquid unlike traditional techniques, which condense on a cold surface.

New physical understanding of heat transfer in BCs has led to low pressure-drop designs. The concept of multi-staging the uniform temperature column in several temperature steps has led to highly effective designs (about 90 per cent). These designs lead to significant cost advantages for CGE over traditional techniques like MVC distillation systems.

The colder fresh water enters the device at the topmost stage and passes through every successive stage in a cross-flow manner (from one end of the stage to the other) and extracts the heat of condensation from the condensing vapor (which heat is in turn used to preheat the feed water).

Each BC stage has a shallow layer (<one inch in height) of this fresh water through which the air/vapour mixture is sparged, creating a multitude of small bubbles. As these bubbles rise through the layer height there is a wake (negative pressure region) created below the bubble which draws in the liquid from the surrounding region. This process sets up a millimetre-size liquid circulation zone which causes continuous renewal of boundary layers formed via the heat and mass transfer phenomena. Such liquid circulations, which cause extremely high levels of turbulence, are created throughout the liquid layer because of the swarm of rising bubbles. The device is designed to be ultra-efficient with optimal bubble size, bubble pitch and liquid layer height that affects bubble residence time, gas velocity and temperature difference per stage.

The newest generation of BC devices have 40 per cent lower heat transfer area requirement than condensers in pure vapour systems like MVC (see Figure 2). At higher salinities (as is the case with the current challenge), MVC uses Grade 5 titanium for heat transfer surfaces. The BC uses the surface of bubbles as heat and mass transfer surfaces as opposed to expensive metallic surfaces used in MVC systems, providing significant capital cost advantages due to lower heat transfer surface area requirements and the use of inexpensive materials.

Thermodynamic balancing

When finite time thermodynamics is used to optimize the energy efficiency of thermal systems, the optimal design is one which produces the minimum entropy within the constraints of the problem (such as fixed size or cost).

This well-established principle, known generally as thermodynamic balancing, was applied to the design of combined heat and mass exchange devices (dehumidifiers and humidifiers) for improving the energy efficiency of CGE systems.

This resulted in novel designs and operating procedures that make it more energy efficient at treating hypersaline wastewater compared to traditional techniques like MVC. At the core of these innovations is a new non-dimensional parameter that was invented to minimize the average local driving force for heat and mass transfer. The fully automated optimization of the system based on this non-dimensional number gives CGE the ability to have constant performance even with varying water quality and volumes.

The physical embodiment in design of the thermodynamic balancing concept is several extraction lines through which specified amounts of the air/vapour mixture is prematurely taken from the humidifier and re-injected at a corresponding location in the dehumidifier (see Figure 1).

In commercial CGE systems, thermodynamic states at different points in the system (at the inlet and outlet and intermediate locations of all unit operations including the humidifier, dehumidifier and heat exchangers) are determined by measuring temperatures, mass flow rates and concentrations.

Using this information, the amount of vapour/gas mixture to be extracted through any given line is determined on a continuous basis by evaluating the optimal operating point which corresponds to a non-dimensional number of 1 for any and all boundary conditions. This novel algorithm is also patented and proprietary to Gradiant.


The decoupling of phase change and heat transfer surfaces is crucial to treating hypersaline water streams because of hardness scaling issues. Scaling tendency increases with increasing salinity; the scale is likely to form on the surface where phase change occurs.

In a MVC system, the phase change and the heat transfer occur on the same surface, resulting in a drop in heat transfer efficiency when scale forms and acts as an insulator.

In CGE, however, the phase change occurs in the humidifier column and the feed water is heated in the heat exchanger. Scale forms on the packing material in the humidifier but does not affect the evaporation or the performance of the system because the carrier gas is in direct contact with the hypersaline water and there is no heat transfer through the material of the packing.

Nevertheless, the packing material only requires cleaning or replacement when the scale build-up severely reduces the fluid flow in the column (see Figure 3). Another benefit of having high tolerance to scaling is that the pretreatment requirements are lower than MVC systems, which further reduces operating costs.

Economic comparison

The CAPEX and OPEX of CGE, MED and MVC are compared for treatment of contaminated FGD wastewater from coal power plants in China for a feed water of 12,500 ppm and a fresh water recovery of 95 per cent (250,000 ppm reject stream) (see Table 1).

The CAPEX of CGE is lower than that of MVC because of the reduced need for expensive metals like Grade 5 titanium. The OPEX of CGE is lower because of lower pretreatment requirements, use of low temperature thermal energy which can be free if solar thermal (evacuated tube) or waste heat is used, and minimal repair and maintenance needs.

In conclusion, the cost effective, high performance treatment of FGD wastewater is becoming mandatory to implement in all coal-fired power plants. Due to the variability of FGD wastewater quality and volumes on top of the high contamination levels, the use of CGE technology is optimal and is significantly more cost effective than other solutions.

Prakash Govindan is Chief Technical Officer at Gradiant Corporation. gradiant.com

Zhifeng Fan is Chief Engineer of Seawater Desalination at Shanghai Electric Company.