Dutch engineering consultant KEMA’s investigation into recovering significant quantities of clean water from flue gas surpassed expectations. Ludwin Daal, process & cooling water consultant at KEMA, discusses the implications of this technological development on the global power industry.
Ludwin Daal, KEMA, The Netherlands
One of the major challenges of this century is the provision of safe drinking water for a growing population. The shortage in water resources in arid areas requires the availability of more efficient and cheaper potable water production processes. In order to make drinking water out of groundwater, it is often sufficient to aerate and disinfect.
However, in large parts of the world, the use of groundwater from aquifers is not possible because of excessive use and global climate change that allow penetration of seawater into the aquifers. Population growth, not surprisingly, leads to greater pollution of aquifers, rendering the water quality unsuitable for drinking water purposes without excessive treatment. In some arid areas there is almost no water available or no water at all.
Interestingly, air can contain large quantities of water vapour. Based on statistics provided by UNESCO, it appears that annually 575 000 km3 of liquid water evaporate and disappear into the atmosphere. This atmospheric water is much more uniformly distributed compared to any other water source. To put this into layman’s terms: If we would evenly cover the earth with this water, we would all be 1.1 metres under water.
In many industrial production processes, evaporated water is present. In these processes, this water vapour is liberated or escapes as ‘waste’ water to the atmosphere. Little development has taken place in capturing the evaporated ‘waste’ water.
Capturing and reusing water from industrial processes can help conserve this valuable resource. An important and localized example of these industrial processes is power generation.
Evaporated water from power generation
A typical 400 MWe coal fired power plant equipped with a flue gas desulphurization (FGD) unit would need about 30 m3 per hour (m3/h) of process water, while at the same time 150 m3/h of water exits the stack in its vapour form.
|Figure 1: This map shows that there are many water stressed areas throughout the world – more than 2.5 billion people live in these areas Source: IPCC (2007)|
Now, if we look at the global installed capacity of approximately 800 GW coal fired and 400 GW gas fired power plants, there is approximately 300 000 m3 water per hour (in liquid form) and 120 000–180 000 m3/h water available respectively in the flue gases that can be targeted for recovery. To put this into perspective, the water consumption in the US is approximately 200 litres per capita per day (l/capita/day) and in arid countries like Ethiopia approximately 20 l/capita/day. From these values it becomes clear that the available water potential could supply many people. Moreover, the total dissolved solids (TDS) level of this evaporated water is much lower compared to surface and/or well water, with the quality of the water captured from a gas fired plant being slightly better compared to that captured in coal fired power plants.
The above calculations do not take into account the evaporated water being emitted by cooling towers, since it is difficult to determine the exact amount and type of cooling towers. That said, open system recirculating cooling towers evaporate between 0.4 m3 and 3.8 m3 water per installed MW/h.
Moreover, aside from energy generating processes, some other industrial processes produce considerable amounts of water vapour as well, for example paper and fibre board producing industries. If only the evaporated water could be captured in an energy efficient way, it could help with the daunting challenge of the global water shortage.
A new idea is born
The idea of capturing water from flue gases did not come out of the blue. In the late 1990s, KEMA conducted extensive work on behalf of the Dutch power industry in transforming surface water into high-quality water. To achieve this, classic technologies like ion-exchange resins, as well as membrane technologies were used.
At the end of the study, it appeared that surface water was a relatively inexpensive source but the quality of the surface water put a lot of stress on the technologies, which made the quest for alternatives relevant. A new possible water supply for the power industry then became interesting: water from their own stack.
A lot of water leaves the stack and its quality is potentially far better than that of the raw surface water.
Thus the first important question to answer was how this water could be captured. Condensation of the water was not an option for several reasons. First, the use of polymer heat exchangers to condense the water out of the flue gas is quite expensive, especially due to the investment costs of hardware.
Secondly, as soon as water starts to condense some of the other compounds present will be captured by the water as well.
Thirdly, considering that flue gas contains only between 8 per cent and 11 per cent water vapour in a power station stack – and up to 20 per cent in a waste-to-energy plant – it does not make sense to cool down the whole stream.
Gas separation membranes were considered to be the answer. In collaboration with the Netherlands’ Twente University, a membrane material was developed that had a high water vapour selectivity over nitrogen.
The need for a high selectivity has two major reasons. Firstly the water has to be as pure as possible, and secondly the technology behind the membranes does not allow any noncondensables. Nitrogen, being the main non-condensable in any flue gas, makes this need for high water vapour selectivity clear.
A basic schematic depiction of how gas separation technology could be implemented in a power plant is shown in Figure 2. In this situation – a coal fired power plant with FGD – the membrane modules are placed downstream of the FGD, but upstream of the flue gas reheater system.
|Figure 2: A schematic of the membrane modules inside a flue gas stream of a coal fired power plant, where captured water vapour is returned to the condenser|
The recovered water vapour (permeate) is transported directly to the existing condenser system, where condensation takes place. From this schematic it is quite clear that the recovered water vapour should not contain any non-condensables, since this will negatively affect reliable operation of the vacuum condenser.
The theoretical considerations and the developed membranes for capturing water from flue gases then had to be tested.
First attempt to capture water with membranes
Several research projects, granted by the Dutch government, have been conducted to test and develop the membrane technology. First, proof-of-principle tests in 2000 showed it was possible to recover water from flue gas by means of gas separation membranes. It appeared that two main issues remained to be tackled: The selectivity of the membrane material and the improvement of the water flux.
A second project, ‘Water and Energy Recovery from Flue Gas’, commenced in 2001 to overcome these issues. The first lab and field experiments in this project were carried out with a hollow fibre membrane module, or ‘air dryer’, which is normally used for water removal from pressurized air.
Economic desk studies showed, however, that using energy to pressurize the feed makes the recovery process financially unviable and consequently not applicable.
Therefore, it was decided to modify the module in such a way that the feed is at the shell side of the module and the permeate – the recovered water vapour – is at the tube side, with an applied vacuum on the tube causing the water vapour to migrate through the membrane surface.
These first results were encouraging, but not yet satisfactory. It appeared that the recovered water quality was not according to expectations, that the water flux was too low and, additionally, that the hollow fibres were too fragile to withstand flue gas conditions.
These last results necessitated further membrane development by Twente University, which resulted in much more robust and selective hollow fibre membranes. The new membranes were then exposed to a series of prolonged field tests.
First prolonged field test
In other tests to simulate normal power plant operation, even with upsets, the membranes were exposed for 24 hours a day, seven days a week for at least 32 weeks. The flue gas was taken from a bypass connected to the stack after the reheating system, a Ljungström gas reheater, and was fed to a membrane chamber containing several hollow, straw-like fibres hand-made by Twente University.
During the test period several disturbances occurred, ranging from relatively high temperatures to fly ash transport caused by malfunctioning of the ESP filter. Despite these events most of the hollow fibres still had a good H2O/N2 selectivity after the test period. The target permeance value was set at >85 000 barrer1, and the exposed fibres showed a selectivity of >100 000 barrer.
However, improvements were still possible. The quality of the captured water still had to be improved if it were to be used as boiler feedwater: a polishing step was necessary in order for it to qualify as ultra-pure water. This was mainly due to sulphate, originating from the residual SO2 in the flue gas, that was dissolved in the water. In addition the average water flux was rather poor at 0.2 litres per square metre per hour (l/m2/h).
The importance of water flux improvement has to do with the fact that the membrane surface area that can be integrated in an existing power plant is limited. Take for example, the 400 MWe coal fired power plant that is required to cover its own demin water needs from captured water.
At an average water flux of 0.2 l/m2/h, an installed membrane surface area of approximately 150 000 m2 is necessary. Interesting figures perhaps for a membrane manufacturer, but quite challenging for a system integrator. For this reason further development to improve the flux was necessary.
The development route
From 2004 to 2010, a combination of field tests inside flue gas streams of a coal fired power plant, a waste-to-energy plant, and a small gas burner resulted in improvements in flux, water quality and also durability of the materials used (see Figure 3).
|Figure 3: Photo’s of one of the field tests at a waste to energy plant. Left the man hole to the duct, and right the membrane modules placed within the flue gas duct|
The development and improvements led to better than expected flux results. A summary of the development route over the years is depicted in Table 1. The test results of the gas burner in 2010 show that, even in ‘dry’ flue gas conditions, water capture is still possible. The results of the last gas burner tests (2009 and 2010) prove that one is able to capture at least 40 per cent of the evaporated water.
|Table 1: General technology development overview. Results are given in water flux, in litres liquid water per m2 membrane area per hour, and water purity determined by the measured specific conductivity|
Energy & water benefits
In conclusion, the ten-year research made clear that the selective hollow fibre membranes are very capable of capturing water from flue gas. At the start of the project, it was assumed it would be possible to recover 20 per cent of the water in flue gas, but research revealed that at least 40 per cent could be recovered, turning a water consuming power plant into a water producer.
Moreover, economic desk studies for a 400 MWe coal fired power plant showed that the water recovery process is competitive with classical IX resin, thermal distillation and membrane technologies, as shown in Figure 4.
Figure 4: Comparison of technologies versus the capture of evaporated water (CapWa) and a commercial tender in the Netherlands in 2009. Savings from reheating and water distribution costs not included Source KEMA, V. Baron (2010)
Here the assumption is made that the price of the water capture membranes is similar to that of reverse osmosis membranes. It should be noted that the graph does not show the savings incurred when flue gas reheating is rendered obsolete. Calculations show that a coal fired power plant can increase its efficiency by about 1 per cent, if reheating is no longer necessary.
The CapWa project
The promising results necessitated further research. Commissioned by the European Union and led by KEMA, 14 partners from Europe, the Middle East and Africa are now working together in an open innovation structure on a follow-up project. This project is CapWa, ‘Capture of Evaporated Water With Novel Membranes’. The project’s main goal is to develop and manufacture membrane modules, ready for industrial use within three to four years.
Within the project, the membrane modules will be demonstrated at gas and coal fired power stations in Spain and Israel, a geothermal well in Tunisia and paper factories in the Netherlands and South Africa. These tests should clear the way for industrial production and large-scale implementation of this new technology. The project started in late 2010, and so far a new membrane module has been designed and membrane research is progressing well.
Gabriel Jinjikashvily, Sustainable R&D manager at the Ruthenberg coal fired power plant, where water is purchased from a water company at a rather high price, says: “This water is fairly clean, but needs two processing steps before we can use it in our processes. However, tests show that the water captured with the membrane technology is so clean that only one step is needed.
“Taking into consideration the water shortage that influences the water price, it would be economically beneficial to implement the water capture technology. Apart from this main benefit to our company, the technology is even more promising thanks to its possible social implications, for example for agricultural purposes.”
Oliver Schuster, vice-president of R&D at Membrana GmbH, said: “If the project turns out to yield positive results, the production of the membranes can be launched very quickly. In fact, the membranes needed for the modules could be industrially produced within a year following the end of the CapWa project. If the demand is very large, however, investments will be needed, in order to upscale the production facilities.”
Reducing the Water Shortage
Aside from the power industry, the technology has the potential to be applied in many different industrial processes. The ultimate choice of utilizing the technology will not only depend on its cost, but also on the industrial process, the local conditions and public opinion. Nevertheless, the industry and, in particular, the power industry can, in just a few years, make a valuable social contribution to the world wide water shortage, making it a better place for those who need it most.
1. 1 barrer = 1 x 10-3 cm-3 (STP)/cm2.sec. cmHg = the amount of gas that permeates through the membrane per unit area, time and unit pressure difference across the membrane.
For more information about the CapWa project, visit www.watercapture.eu
Power Engineering International Archives
View Power Generation Articles on PennEnergy.com