Mikko Rantanen, Wärtsilä, Finland
The terms ‘water use’ and ‘water consumption’ are both used in connection with power plants, especially when talking about the requirements of cooling systems. In the cooling process of a power plant, water can be either used or consumed, whereas in other power production related operations water is normally consumed. Water ‘use’ indicates that the same volume of water that is taken from the source into the plant is directed back to that source. Water use takes place, for example, in once-through cooling systems, where after being used for cooling, the water is directly returned to the source.
Water ‘consumption’, on the other hand, means that a certain amount of water is actually consumed in the operation. In cooling applications, water consumption indicates that only part of the water used for cooling is directed back to the water source, with the remainder being lost through evaporation and drift during the cooling process.
Decentralized plants using radiator cooling can be located in areas with limited water sources available.
The scale of water consumption in a power plant is very much dependent upon the cooling method applied. Sanitary use consumption of course relates to the number of people working on site, which again depends on the type and location of the plant. For example, a gas engine plant located in a developed country normally requires only a few operators, and thus consumes very little water for sanitary use. Conversely, the same type of plant in a developing country usually has many more workers, which naturally results in higher water consumption.
The level of water consumption through auxiliary processes depends on the fuel and plant type. Water consumption in gas and LFO (light fuel oil) plants is lower than in HFO (heavy fuel oil) plants, since no fuel treatment is needed. In addition to sanitary and auxiliary process water consumption, local operating methods and habits also play a role. Water consumption in operations such as area cleaning and parts washing can, for example, vary remarkably from plant to plant.
However, all of these water-consuming factors become insignificant within the larger picture if the cooling method applied in the plant is of a water consuming type. The main cooling methods in power production applications are once-through cooling, tower cooling, and closed-loop radiator cooling.
The flow of water can be very high due to the temperature limitations of the discharged water. When the permitted temperature increase at the discharge point is limited, the water flow is often much greater than actually needed for the plant’s cooling system.
This will result in pumping and investment costs for the cooling water intake and discharge systems being higher than necessary. The discharged water temperature may have limitations since a warm water discharge may have an adverse impact on the local ecosystem in the receiving water.
The majority of the cooling tower makeup water is evaporated. To be able to keep the water quality in the cooling tower circuit acceptable, part of the cooling tower water has to be discharged as ‘bleed-off ‘, which has a far greater concentration of impurities than in the cooling tower make-up water.
For proper operation, the cooling tower needs a continuous dosing of chemicals to prevent scaling and fouling within the cooling circuit. Limitations on the effluent composition affect the selection of chemicals since they are also concentrated into the bleed off water.
Here too, the warm and saline water discharge may have an adverse impact on the local ecosystem within the receiving water.
In a closed circuit cooling system, water consumption is negligible. The heat is transferred to the airflow forced through the radiators by electrical fans. The electrical consumption of the fans is reduced by a frequency converter when the ambient temperature is lower.
Once-through cooling systems are typically applied in large-scale centralized power plants using steam turbines, and are rarely employed in decentralized engine and gas turbine installations. In Wärtsilä power plants, dry cooling by radiators is the most common solution.
In engine power plants, cooling is necessary in order to maintain the charge air, lube oil and jacket cooling water temperatures at the required levels. With gas turbines, on the other hand, the above mentioned cooling methods are used for intercooling, or for secondary cooling of the mechanical refrigeration and absorption chiller systems (used for inlet-air cooling).
Regardless of the power plant size and type, water use and consumption always have both economic and environmental impacts, especially when groundwater is used for cooling in drought sensitive areas.
While large, centralized plants using once-through cooling systems have to be located near water sources, decentralized plants using radiator cooling have a very moderate water usage and can be located in areas with limited available water sources.
Water consumption in power plants
Water consumption in Wärtsilä power plants is low because the cooling is normally via air-cooled radiators. In some rare cases cooling towers may be used and this will naturally increase the water consumption, but even in this kind of installation water consumption is relatively low.
In the European Commission’s ‘Reference Document on Best Available Techniques for Large Combustion Plants’ the water consumption of an engine power plant has been estimated. According to this document “an engine driven power plant usually preserves its water”.
However, consumption in a Wärtsilä power plant is even lower than the document states. For example, for a 130 MW power plant, the same size as that referenced in the document, water consumption is around 150 m3/h using a cooling tower, whereas in the document it has been estimated to be 220 m3/h.
As a comparison, the document gives a value of 500 m3/h for a steam turbine plant of the same size. In the case of radiator cooling, the water consumption is negligible in comparison to these figures.
The estimated total water consumption, including all the water needed in the plant’s operations, is around 9 l/MWh for a Wärtsilä HFO plant, and around 2.7 l/MWh for a LFO and gas plant. These values are based on the assumption that cooling is via radiators, which is the standard solution.
In HFO plants, water is consumed by the process water, boiler water and sanitary water. The process water consumption includes water used by the fuel oil and lube oil separators, turbo washing, the oily water treatment system and workshop operations.
Some minor amounts may also be used in engine cooling and by evaporation. Another consumer group are boilers. HFO plants have their own consumption boilers that are used to produce steam for plant operation purposes.
The boilers consume water in the form of make-up water, and also in cooling blow-down from the boiler. The third and smallest consumer is sanitary use.
As opposed to HFO plants, in LFO and gas plants the majority of the water is consumed by the sanitary system. The process water consumption in these plants mainly consists of that used in workshop operations. As with HFO plants, engine cooling may also consume some minor amounts through leakages and evaporation.
However, cooling is not the main factor in the water consumption of either type of plant, unlike in many other power plant types. A principal water flow diagram for a Wärtsilä HFO plant is shown in Figure 1.
Figure 1. Principal water flow diagram of an HFO plant
Comparison to gas turbine plants
Wärtsilä power plants are well known for their excellent performance, even in difficult conditions, such as high ambient temperatures or altitudes. The performance of the plant remains high regardless of whether the plant is installed in tropical or arctic conditions, or if the plant is located at high altitude.
Moreover, the installation’s location has no significant effect on the plant’s water consumption. Gas turbine performance is normally given in ISO conditions, meaning a dry bulb temperature of 15 °C, a relative humidity of 60 per cent, and an atmospheric pressure of 1 bar (sea level). However, most gas turbine installations are not operating in such conditions.
The performance of a gas turbine decreases the more actual conditions differ from these norms. Comparisons of the derating factors because of site conditions between a Wärtsilä gas engine and a competing gas turbine plant are shown in Figure 2.
Figure 2. Comparisons of derating factors due to external conditions between a Wärtsilä gas engine plant and a competing gas turbine plant.
Changes due to local conditions do not have a remarkable effect on the derating factor of a Wärtsilä engine. However, at high altitudes gas turbines lose their performance significantly. The same thing occurs in hot conditions, but this can be prevented partly by inlet-air cooling.
Importance of inlet-air cooling
Several methods are used for inlet-air cooling in gas turbine plants. Possible options include media-type evaporative coolers, fogging systems, mechanical refrigeration systems, and absorption chillers.
Of these methods, mechanical refrigeration systems and absorption chillers are the most effective because their function is not limited by the ambient wet-bulb temperature. However, they have high initial capital costs, high operations and maintenance (O&M) costs, and relatively long delivery and installation times.
Moreover, they require expertise to operate and maintain the plant. As an example, the initial capital cost of chillers is in the magnitude of $1 million for a gas turbine with an output of 41 MW ISO.
The preferred solutions are in many cases media-type evaporative coolers and fogging systems. This is because of their relatively low capital and O&M costs, quick delivery and installation times, and easier operation.
A media-type evaporative cooler consists of a wetted honeycomb-like medium, through which the inlet air is pulled. As the air flows through the medium, water is evaporated from the surfaces. The evaporation naturally requires energy, and thus the inlet air is cooled.
The fogging system is also based on cooling air through the evaporation of water, but instead of using a medium, the water is atomized into fog-droplets. The media-type evaporative cooler is the most widely used cooling method, and the fogging system the second most frequent one.
However, these methods also have drawbacks; the limitation on capacity improvement, and a high dependence of performance on the ambient wetbulb temperature. And last but not least, both these methods consume relatively large quantities of water.
For fogging systems demineralized water is always needed, whereas for evaporative coolers, the requirements on water quality are less stringent, although bleed off is needed to remove concentrated impurities.
A fogging system supplier has published typical water consumption values for 11 °C inlet air-cooling. When the ambient temperature is 38 °C and the wet-bulb temperature 20 °C, the water consumption for an industrial gas turbine with 41 MW ISO output is 30.3 l/min.
Without the inlet air cooling the output would be as low as 25 310 kW, whereas with cooling it can be increased by 8165 kW (32.3 per cent) to 33 475 kW.
However, as the figures show, even with cooling the output remains clearly under the ISO value. The water consumption of a media-type evaporative cooler is in the same magnitude.
The water consumption of 30.3 l/ min means around 1800 l/h. In the form of specific consumption, the value is 54 l/MWh. As mentioned earlier, the estimated total water consumption for a Wärtsilä HFO plant is around 9 l/MWh, and for a LFO and gas plant, 2.7 l/MWh.
A comparison of these values shows that the water consumption in Wärtsilä engine plants is significantly lower than in a gas turbine installation using a wet cooling method.
Moreover, the consumption figure in the gas turbine installation is only for inlet-air cooling, and is thus missing other consumers, such as sanitary facilities, the workshop, washing water, etc.
In conclusion, when compared to gas turbine installations, the main benefits of Wärtsilä engine plants are their high tolerance to extreme conditions, and their low water consumption regardless of the prevailing conditions. The more conditions deviate from ‘standard’, the greater the difficulty is for competing gas turbine plants to maintain output of the plant at a satisfactory level.
To minimize the drop in output, they are obliged to use inlet-air cooling systems. Such systems are either both expensive and difficult to operate, or they consume significant volumes of water. It seems that in gas turbine applications, water consumption is closely related to the economical optimization of the plants.
Currently, water consuming cooling methods are more common because of their lower initial capital and O&M costs. High water consumption anyhow makes them environmentally questionable. In this regard, Wärtsilä power plants offer a big advantage.
Because of their low water consumption, the plants can be operated in locations with restricted water supplies and will, irrespective of the cooling solution, have a higher electrical efficiency.
Because of their low water consumption, Wärtsilä plants have a minimal discharge of wastewater. Another factor is their low usage of water treatment chemicals, which therefore means only a minor risk of chemical spillages.
The amount of wastewater produced in a HFO plant is approximately 4 l/MWh. This process wastewater, usually called oily wastewater, is produced for example, in fuel and lube oil separators, and in plant area washing.
The oily wastewater is treated within the plant to comply with World Bank Guidelines for Thermal Power before being discharged.
The separated oil is collected in a sludge tank and then utilized or disposed of in an environmentally sound way. In a gas plant there is basically no process wastewater produced.
In many cases there is no high quality water available and the water needed is produced locally by taking the water from the nearest water source and treating it to fit the criteria for power plant use.
Most commonly, the process water treatment system supplied by Wärtsilä is built in an insulated and air conditioned container. The system can then be factory-assembled and tested, thus minimizing the onsite installation time and thereby both reducing the commissioning time, and ensuring the high quality of the work.
The normal treatment process includes sand filtration, ion exchange softening, and reverse osmosis. Optionally, the system can be equipped with raw water chlorination, activated carbon filtration, and anti-scalant dosing for the reverse osmosis.
Experience has shown that these selected processes can cope with a wide variety of incoming raw water with relatively small modifications. MEE